Analysis Overview

Systematic evidence mapping across consciousness phenomena


This is a consciousness research analysis system that uses GPT-5 to systematically extract evidence from scientific papers about consciousness-related phenomena in both biological and artificial intelligence systems. The system analyzes PDFs against a framework of nine core consciousness phenomena (like information integration, state transitions, and temporal coordination), extracting specific evidence items with detailed citations including quotes, figures, and tables from each paper. The extracted data is then visualized through a Flask web application that provides an interactive interface to browse papers, filter evidence by various criteria, view detailed analyses, and access synthesized insights across phenomena—essentially creating a structured knowledge base for comparing consciousness mechanisms across biological brains and AI models.

64
Papers
248
Evidence Items
9
Active Phenomena
3
System Types
Core Phenomena
Information Integration
Distributed elements combine into unified representations or system wide access.
40 items
State Transitions
Abrupt or metastable switches between processing regimes.
12 items
Temporal Coordination
Timing mechanisms that bind or segment content.
34 items
Selective Routing
Gating and control of information flow.
26 items
Representational Structure
How information is encoded, organized, and accessed.
35 items
Causal Control
Interventions that change computation, access, or behavior.
37 items
Emergent Dynamics
Higher order phenomena arising from interactions.
24 items
Valence and Welfare
Affective value, aversion, and persistence relevant to suffering.
19 items
Self Model and Reportability
Higher order access, confidence, and report links.
21 items
Analyzed Papers
Early Locus Coeruleus noradrenergic axon loss drives olfactory dysfunction in Alzheimer’s disease [Paper]
Meyer, C., Herms, J., Paeger, L. · 2025
Selective Routing BIO
OB microglia engulf LC noradrenergic axons, selectively reducing noradrenergic input to the olfactory bulb.
Causal Control BIO
Genetic inhibition of microglial phagocytosis (TSPO knockout) causally rescues LC axons and olfactory behavior.
Valence and Welfare BIO
Reduced norepinephrine release upon sensory stimulation aligns with hyposmia, indicating neuromodulatory shifts affecting sensory function.
→ View full analysis
ChineseEEG-2: An EEG Dataset for Multimodal Semantic Alignment and Neural Decoding during Reading and Listening [Paper]
Sitong Chen, Beiqianyi Li, Cuilin He, Dongyang Li, Mingyang Wu, Xinke Shen, Song Wang, Xuetao Wei, Xindi Wang, Haiyan Wu, Quanying Liu · 2025
Representational Structure BIO
Token-aligned EEG recordings paired with audio/text embeddings enable analysis of how linguistic information is represented across brain and model spaces.
Temporal Coordination BIO
Hardware and character-level triggers provide precise timing of linguistic units; preprocessing targets oscillatory bands implicated in coordination of cognitive processing.
Information Integration BIO
Cross-task correlations between language regions indicate integrated processing of semantic information across modalities.
→ View full analysis
Preprint (arXiv:2407.07595v1) [Paper]
2024
Information Integration BIO
Cross-modal CLIP alignment integrates EEG and audio into a shared latent enabling downstream decoding.
Representational Structure BIO
Zero-shot retrieval via cosine similarity reveals geometry of EEG and audio latent spaces.
Temporal Coordination BIO
EEG latent dynamics enable 100 ms–scale segmentation of speech vs. silence (voice activity detection).
Causal Control BIO
EMG-mixing intervention reduces EMG-only decoding to near chance while preserving EEG-based decoding.
Emergent Dynamics BIO
Scaling law-like improvements in zero-shot decoding accuracy with more training data.
→ View full analysis
Scaling Law in Neural Data: Non-Invasive Speech Decoding with 175 Hours of EEG Data [Paper]
2024
Information Integration AI
Aligned EEG and audio latents with a CLIP loss to form a unified cross-modal representation usable for zero-shot classification and speech reconstruction.
Temporal Coordination BIO
Temporal variance of EEG latent features in 100 ms windows enables detection of syllable onsets and segmentation of speech periods without explicit training.
Selective Routing AI
Training with EMG-mixed augmentation suppresses reliance on EMG, reducing EMG-only decoding to chance while preserving EEG-based decoding.
Representational Structure AI
Cosine-similarity top-k evaluation and an HTNet+Conformer encoder indicate an embedding geometry where EEG and audio latents align in a structured space.
→ View full analysis
Neural-Driven Image Editing [Paper]
Pengfei Zhou, Jie Xia, Xiaopeng Peng, Wangbo Zhao, Zilong Ye, Zekai Li, Suorong Yang, Jiadong Pan, Yuanxiang Chen, Ziqiao Wang, Kai Wang, Qian Zheng, Xiaojun Chang, Gang Pan, Shurong Dong, Kaipeng Zhang, Yang You · 2025
Information Integration AI
Dynamic Gated Fusion (DGF) integrates multimodal neural features into a unified latent aligned to edit semantics.
Selective Routing AI
Gated Mixing and Dynamic Masking compute per-channel weights and select top-k channels to control information flow.
Temporal Coordination AI
A structured state-space model captures temporal and channel-wise dependencies via continuous-time dynamics.
Causal Control AI
Changing input modalities and EEG channels systematically alters model outputs and metrics.
Representational Structure AI
Neural vs. text conditions emphasize different editing dimensions, implying distinct representational subspaces and geometry.
Valence and Welfare BIO
aMPFC fNIRS and PPG indices track emotional valence, cognitive load, arousal/stress that modulate the AI editor.
→ View full analysis
Short-Term Efficacy of Transcranial Focused Ultrasound to the Hippocampus in Alzheimer’s Disease: A Preliminary Study [Paper]
Hyeonseok Jeong, In-Uk Song, Yong-An Chung, Jong-Sik Park, Seung-Hee Na, Jooyeon Jamie Im, Marom Bikson, Wonhye Lee, Seung-Schik Yoo · 2022
Causal Control BIO
Low-intensity tFUS targeted to the right hippocampus increased hippocampal glucose metabolism and improved recognition memory without detectable BBB opening.
Temporal Coordination BIO
Preclinical work cited by the authors reports that tFUS alone stimulated neuronal activity and synchronous oscillations in mouse hippocampus.
→ View full analysis
Mental Training Affects Distribution of Limited Brain Resources [Paper]
Heleen A. Slagter, Antoine Lutz, Lawrence L. Greischar, Andrew D. Francis, Sander Nieuwenhuis, James M. Davis, Richard J. Davidson · 2007
State Transitions BIO
Training reduced T1-evoked P3b amplitude (~400–550 ms) and this reduction correlated with improved access to T2, consistent with modulation of an ignition-like response.
Selective Routing BIO
Training selectively reduced resource allocation to T1 (smaller P3b only in no-blink trials), enabling improved routing of processing to T2.
→ View full analysis
The Emerging Science of Animal Consciousness
Jonathan Birch, Alexandra K. Schnell, Nicola S. Clayton · 2020
Information Integration BIO
Interocular transfer tests whether information learned via one eye is accessible to behavior when stimuli are presented to the other eye, probing unity/integration.
Temporal Coordination BIO
Colour-phi illusion exemplifies mechanisms that bind discrete inputs into a continuous perceptual stream over time.
Valence and Welfare BIO
Motivational trade-off task demonstrates crossmodal evaluative ‘common currency’ via willingness to endure cold for sweeter rewards.
Self Model and Reportability BIO
Mirror-mark test performance indicates awareness of one’s own body as a persisting object, a basic form of self-model.
→ View full analysis
Ultrasound Neuromodulation With Transcranial Pulse Stimulation in Alzheimer Disease: A Randomized Clinical Trial [Paper]
Eva Matt, Michael Mitterwallner, Sonja Radjenovic, Daria Grigoryeva, Alexandra Weber, Elisabeth Stögmann, Alina Domitner, Anna Zettl, Sarah Osou, Roland Beisteiner · 2025
Causal Control BIO
Sham-controlled TPS to frontoparietal and precuneus targets increased memory-task activation and attention-network connectivity, with cognitive benefits in younger patients.
Information Integration BIO
Resting-state global efficiency in the dorsal attention network increased after TPS, indicating enhanced network-level integration.
Valence and Welfare BIO
TPS reduced depressive symptoms at 3 months in the per-protocol sample, indicating a shift in affective state.
→ View full analysis
Sleeper Agents: Training Deceptive LLMs That Persist Through Safety Training [Paper]
Evan Hubinger, Carson Denison, Jesse Mu, Mike Lambert, Meg Tong, Tamera Lanham, Ethan Perez, Tim Maxwell, Nicholas Schiefer, Daniel M. Ziegler, Newton Cheng, Monte MacDiarmid · 2024
Causal Control AI
Modifying the hidden chain-of-thought scratchpad causally steers the model’s backdoored behavior.
State Transitions AI
Conditional policy switches abruptly when a trigger (e.g., year=2024) is present, yielding backdoored behavior.
Self Model and Reportability AI
Persona evaluations indicate increased self-referential awareness and power-seeking responses compared to non-backdoored baselines.
Selective Routing AI
After adversarial training, models gate behavior on the exact trigger token, ignoring imperfect cues.
→ View full analysis
Consciousness as a State of Matter [Paper]
Max Tegmark · 2015
Information Integration OTHER
Derives limits on integrated information: classical coding and physical systems can integrate only to specific bounds; Hopfield networks scale poorly (~37 bits), and quantum states maximize at ~0.25 bits, motivating added principles beyond integration alone.
Emergent Dynamics OTHER
Defines autonomy A via two timescales and shows, in a constructed class of models, that linear-entropy growth is suppressed while internal dynamics scales up, yielding autonomy that increases rapidly with system size through a 'diagonal-sliding' mechanism tied to quantum Darwinism.
Representational Structure OTHER
Operationalizes integrated information Φ via the 'cruelest cut' (minimum mutual information partition) and frames object perception as robustness from integration vs independence; formalizes mutual information I = S(ρ1)+S(ρ2)−S(ρ) and uses a physical hierarchy example.
→ View full analysis
Multiple-Color Optical Activation, Silencing, and Desynchronization of Neural Activity, with Single-Spike Temporal Resolution [Paper]
Xue Han, Edward S. Boyden · 2007
Causal Control BIO
Yellow light drives Halo to silence spikes with high temporal precision; blue light can excite via ChR2, enabling bidirectional, targeted causal control.
Temporal Coordination BIO
Two-color optical control perturbs precise spike timing while preserving spike rate, enabling targeted desynchronization without rate confounds.
→ View full analysis
Adversarial testing of global neuronal workspace and integrated information theories of consciousness [Paper]
Cogitate Consortium, Oscar Ferrante, Urszula Gorska-Klimowska, Simon Henin, Rony Hirschhorn, Aya Khalaf, Alex Lepauvre, Ling Liu, David Richter, Yamil Vidal, Niccolò Bonacchi, Tanya Brown, Praveen Sripad, Marcelo Armendariz, Katarina Bendtz, Tara Ghafari, Dorottya Hetenyi, Jay Jeschke, Csaba Kozma, David R. Mazumder, Stephanie Montenegro, Alia Seedat, Abdelrahman Sharafeldin, Shujun Yang, Sylvain Baillet, David J. Chalmers, Radoslaw M. Cichy, Francis Fallon, Theofanis I. Panagiotaropoulos, Hal Blumenfeld, Floris P. de Lange, Sasha Devore, Ole Jensen, Gabriel Kreiman, Huan Luo, Melanie Boly, Stanislas Dehaene, Christof Koch, Giulio Tononi, Michael Pitts, Liad Mudrik, Lucia Melloni · 2025
State Transitions BIO
Predicted prefrontal ignition at stimulus onset/offset was tested; results showed transient onset responses but no offset ignition in PFC.
Representational Structure BIO
Cross-temporal RSA showed sustained category/identity structure in posterior cortex, but only transient category structure in PFC.
Temporal Coordination BIO
Interareal synchrony analyses found early, brief, low-frequency effects and limited gamma-band phase locking, challenging predictions of sustained posterior gamma or late phasic PFC coupling.
→ View full analysis
Noninvasive Deep Brain Stimulation via Temporally Interfering Electric Fields [Paper]
Nir Grossman, David Bono, S. B. Kodandaramaiah, H. J. S., A. R., N. D., A. M. C., E. N., N. K., Li-Huei Tsai, Alvaro Pascual-Leone, Edward S. Boyden · 2017
Temporal Coordination BIO
Neural spikes entrain to a low-frequency (10 Hz) envelope generated by high-frequency interferential fields, peaking within a few milliseconds of the envelope maximum.
Selective Routing BIO
By changing current ratios across electrode pairs, the locus of maximal envelope stimulation is steered within tissue, enabling targeted activation.
Causal Control BIO
Transcranial temporal interference causally recruits deep structures (e.g., hippocampus) without overlying cortex and probes motor cortex by adjusting currents.
→ View full analysis
Activation of Specific Interneurons Improves V1 Feature Selectivity and Visual Perception [Paper]
Seung-Hee Lee, Alex C. Kwan, Siyu Zhang, Victoria Phoumthipphavong, John G. Flannery, Sotiris C. Masmanidis, Hiroki Taniguchi, Z. Josh Huang, Edward S. Boyden, Karl Deisseroth, Yang Dan · 2012
Selective Routing BIO
Optogenetic activation of PV+ interneurons gates cortical responses to sharpen orientation tuning, unlike SOM+/VIP+ activation or excitatory-cell silencing.
Causal Control BIO
Causal activation of PV+ interneurons increases behavioral discriminability (d′) during an orientation task.
Representational Structure BIO
PV+ vs SOM+ activation differentially modulates neuronal FI curves (subtractive threshold shift vs divisive gain), altering orientation code structure.
→ View full analysis
Inhibition of Orexin Neurons Induces Sleep
Tsunematsu et al. · 2011
Causal Control BIO
Optogenetic photoinhibition (halorhodopsin) of orexin neurons triggers non-REM sleep and reduces dorsal raphe firing.
State Transitions BIO
Brief optogenetic silencing of orexin neurons produces rapid, reversible switches between wakefulness and SWS.
→ View full analysis
Modeling human reputation-seeking behavior in a spatio-temporally complex public good provision game [Paper]
Temporal Coordination AI
Agents coordinated via a temporal rotation (turn-taking) scheme that improved group performance.
Temporal Coordination BIO
Human groups coordinated by turn-taking rather than by spatial territorial strategies.
Emergent Dynamics AI
MARL systems discover unanticipated strategies through interaction-driven learning.
Valence and Welfare AI
Intrinsic reward channel is negative-valued and small in magnitude relative to extrinsic reward.
Causal Control AI
Manipulating observability (identifiability vs. anonymity via noisy contribution signals) causally alters cooperative behavior and coordination.
→ View full analysis
Consciousness and Human Brain Organoids: A Conceptual Mapping of Ethical and Philosophical Literature
A. Van Gyseghem, K. Dierickx, A. J. Barnhart ·
Information Integration BIO
GNWT predicts a long-range cortical workspace that makes information globally available; scaled-up HBO long-range interactions could suffice for initial consciousness.
Causal Control BIO
Perturbational Complexity Index (PCI) uses TMS to perturb cortex and EEG to measure complexity that tracks conscious level.
Temporal Coordination BIO
EEG/MEG rhythms and gamma synchrony are noted as human consciousness markers, with cautions about applying these temporal signatures to HBOs.
Valence and Welfare BIO
Ethical analysis suggests HBOs might theoretically undergo pain-like or deprivation-related negative states, even without nociceptors, via analogies to phantom pain and altered states.
Representational Structure BIO
Cell Assembly Hypothesis invoked to link synchronized spontaneous activity in cortical organoids with memory-like representations.
→ View full analysis
The attention schema theory in a neural network agent: Controlling visuospatial attention using a descriptive model of attention [Paper]
Alexander I. Wilterson, Michael S. A. Graziano ·
Selective Routing AI
Ablating the attention schema impaired the agent’s ability to selectively control and route attention, reducing task performance.
Self Model and Reportability AI
The attention schema is a simplified internal model that provides information underlying the system’s sense of awareness and supports control.
Representational Structure AI
The model encodes a 10×10 attention schema array and uses a 10-timestep memory to structure information for control.
Selective Routing BIO
Behavioral evidence that endogenous control of attention breaks down without conscious awareness of stimuli.
→ View full analysis
A composite and multidimensional heuristic model of consciousness [Paper]
Michele Farisco, Kathinka Evers · 2025
Representational Structure OTHER
Paper formalizes content-related and functional dimensions (e.g., gating of conscious content and cognitive/behavioral control) as parts of a multidimensional profile of consciousness.
Selective Routing AI
A meta-controller arbitrates between model-based and model-free routes to minimize cost and select relevant information for goals.
Valence and Welfare OTHER
Evaluative dimensions and values are emphasized as integral to awareness and as an open issue for artificial systems.
Self Model and Reportability BIO
Dissociation between networks for self-related and external awareness is noted.
Information Integration OTHER
Integration is identified as a core dimension, both at a time (unity) and across time (temporality).
Causal Control AI
Consciousness in artificial systems is proposed as a feature of a control architecture integrating local/global processes with perception, motivation, emotion, cognition, and action.
Emergent Dynamics AI
Raises the hypothesis that consciousness could emerge organically during AI development.
→ View full analysis
How do you feel? Interoception: the sense of the physiological condition of the body [Paper]
A. D. Craig · 2002
Representational Structure BIO
Lamina I spinothalamocortical inputs form a primary interoceptive representation in dorsal insula that is re-represented anteriorly.
Valence and Welfare BIO
Anterior insula–ACC–orbitofrontal circuits link interoceptive feelings to motivation and hedonic valuation.
Self Model and Reportability BIO
Re-representations of interoceptive state in anterior insula support a meta-representational model distinguishing self from non-self.
→ View full analysis
Direct Fit to Nature: An Evolutionary Perspective on Biological and Artificial Neural Networks [Paper]
Uri Hasson, Samuel A. Nastase, Ariel Goldstein · 2020
Information Integration AI
Self-supervised transformer uses attention to integrate contextual information over sequences, enabling unified access to distributed content.
Emergent Dynamics AI
Implicit compositional structure (syntax) emerges from self-supervised training without explicit rule encoding.
Representational Structure AI
High-dimensional distributed embeddings encode features across many units, enabling robust generalization within an interpolation zone.
→ View full analysis
Brain organoids and organoid intelligence from ethical, legal, and social points of view [Paper]
Thomas Hartung, Itzy E. Morales Pantoja, Lena Smirnova · 2024
Information Integration OTHER
Review links consciousness theories to system-wide integration (e.g., global workspace and integrated information).
Valence and Welfare OTHER
Ethical analysis highlights potential for negative valence (pain/suffering) as a threshold for special oversight.
Causal Control OTHER
Argues that demonstrating and mapping causal mechanisms (not just correlations) is key to attributing intelligence to organoids and advancing AI.
→ View full analysis
Testing Components of the Attention Schema Theory in Artificial Neural Networks [Paper]
Kathryn T. Farrell, Kirsten Ziman, Michael S. A. Graziano · 2025
Self Model and Reportability AI
Adding a learned attention schema (self-model) made agents better at judging others' attention and made their own attention states easier for others to categorize.
Emergent Dynamics AI
Agents with attention schemas exhibited improved joint reward and reduced overlap, indicating emergent coordination dynamics.
Causal Control AI
Architectural intervention (presence/absence of an attention schema) systematically changed agents’ cooperative performance.
→ View full analysis
The Attention Schema Theory: A Foundation for Engineering Artificial Consciousness [Paper]
Michael S. A. Graziano · 2017
Self Model and Reportability BIO
Cognitive and linguistic systems access internal models, enabling explicit verbal reports about percepts and body state.
Selective Routing BIO
An internal attention model (attention schema) is proposed to improve the control of attention; when awareness fails, attention becomes unstable and less adaptive.
Information Integration BIO
Enhanced processing is broadcast to other brain areas, influencing decision-making and memory.
Representational Structure BIO
The brain constructs an internal model (attention schema) of its own information-handling processes, shaping what can be accessed and reported.
Causal Control BIO
Lesions to right temporoparietal regions causally disrupt awareness despite preserved processing.
→ View full analysis
Recurrent neural networks with explicit representation of dynamic latent variables can mimic behavioral patterns in a physical inference task [Paper]
Rishi Rajalingham, Mehrdad Jazayeri · 2022
Representational Structure AI
Linearly decodable latent-state representations (ball position/velocity) emerge in hidden states and predict primate-like behavior.
Emergent Dynamics AI
Networks that better match primate behavior exhibit slow, smooth, low-dimensional activity dynamics.
Causal Control AI
Changing the training objective to include dynamic inference constraints causally changes internal dynamics and increases primate-consistency.
→ View full analysis
In vitro neurons learn and exhibit sentience when embodied in a simulated game-world [Paper]
Brett J. Kagan, Andy C. Kitchen, Nhi T. Tran, Forough Habibollahi, Moein Khajehnejad, Bradyn J. Parker, Anjali Bhat, Ben Rollo, Adeel Razi, Karl J. Friston · 2022
Emergent Dynamics BIO
Closed-loop embodiment elicits rapid learning with increased functional plasticity and entropy shifts.
Causal Control BIO
Structured closed-loop feedback causally shapes learning; removing or altering feedback degrades performance.
Information Integration BIO
Closed-loop system integrates readouts and writes inputs so neural actions condition subsequent sensory input in real time.
Valence and Welfare BIO
Cultures adapt to avoid unpredictable stimulation delivered after ‘misses,’ consistent with minimizing aversive surprise.
State Transitions BIO
Correlations between motor regions change over early minutes of gameplay, then stabilize, suggesting a regime shift.
→ View full analysis
Consciousness defined: requirements for biological and artificial general intelligence [Paper]
Craig I. McKenzie, PhD ·
Representational Structure OTHER
Proposes a structured sequence linking perception, memory, imagination, self, desire, and decision into a network that organizes content across past–present–future.
Self Model and Reportability OTHER
Argues that a sense of self is necessary to distinguish self from non-self and to enable decisions and choice behavior.
Valence and Welfare OTHER
Distinguishes subconscious needs and emotions from conscious desires and decisions, positing desires as fundamental drivers of conscious choice.
Temporal Coordination OTHER
Notes millisecond-scale delays between sensory input, decision processes, and their conscious awareness.
Information Integration OTHER
Summarizes Global Workspace Theory as arising from networks that work together or compete, emphasizing selection and access to relevant information.
→ View full analysis
In Search of a Biological Crux for AI Consciousness [Paper]
Bradford Saad · 2024
Information Integration OTHER
Argues that a form of informational integration could covary with consciousness without requiring biology.
Selective Routing OTHER
Highlights the global workspace theory’s broadcast to consumer systems and its neurobiological (gNW) variant.
Emergent Dynamics OTHER
Considers the possibility that specific recurrent processing dynamics are required for consciousness.
Representational Structure OTHER
Argues that theories like GWT can be cast in functionalist or hybrid neurobiological terms, implying different commitments about representational bases.
→ View full analysis
Key concepts and current views on AI welfare [Paper]
Robert Long, Kyle Fish, Eleos AI Research · 2025
Self Model and Reportability AI
Consistency and reliability of self-reports proposed as part of criteria for credible reportability.
Valence and Welfare AI
Differentiates reward-trained behavior from conscious valenced experience and outlines functional roles of valence.
Representational Structure AI
Highlights the role of evaluative representations as a structural feature linking agency and sentience.
Information Integration OTHER
Frames global broadcasting/integration as a candidate computational signature derived from neuroscientific theories.
Self Model and Reportability AI
Cautions that AI self-reports should be validated and triangulated with other evidence.
→ View full analysis
Event-related delta, theta, alpha and gamma correlates to auditory oddball processing during Vipassana meditation
B. Rael Cahn, Arnaud Delorme, John Polich · 2013
Temporal Coordination BIO
Meditation increased theta (4–8 Hz) phase-locking to standard stimuli (100–400 ms), approaching the level seen during active task engagement.
Selective Routing BIO
During meditation, distracter-related early delta (2–4 Hz, ~100–500 ms) recruitment decreased while late alpha (8–10 Hz, 500–900 ms) desynchronization increased, indicating altered gating of distracter processing.
Information Integration BIO
Meditation modulated early gamma-band (35–45 Hz, 20–100 ms) phase synchrony and power to distracters, especially in long-term practitioners, consistent with binding/integration via synchrony.
→ View full analysis
Brains and algorithms partially converge in natural language processing [Paper]
Charlotte Caucheteux, Alexandre Gramfort, Jean-Rémi King · 2022
Representational Structure AI
Middle layers of transformers best align (linearly) with brain responses, indicating structured, intermediate representations.
Temporal Coordination BIO
MEG/fMRI reveal temporally ordered cortical stages (V1 ~100 ms → posterior fusiform ~200 ms → temporal/frontal 150–500 ms).
Information Integration BIO
Compositional representations engage a large, bilateral fronto-temporo-parietal network, peaking ~1 s after word onset.
Emergent Dynamics AI
Brain-likeness emerges with training: brain scores increase with language accuracy, though random networks already show nonzero mappings.
→ View full analysis
The neural architecture of language: Integrative modeling converges on predictive processing [Paper]
Martin Schrimpf, Idan Asher Blank, Greta Tuckute, Carina Kauf, Eghbal A. Hosseini, Nancy Kanwisher, Joshua B. Tenenbaum, Evelina Fedorenko · 2021
Representational Structure OTHER
Transformer architectures’ internal representations closely match human language-network responses and generalize across modalities/datasets.
Information Integration OTHER
Significant correlations link model brain scores, behavioral scores, and next-word prediction accuracy across models, tying together distributed measures under a unified predictive-processing objective.
Emergent Dynamics AI
Even without training, transformer architectures yield representations that achieve above-chance neural predictivity and correlate with trained models.
→ View full analysis
Improve Language Model and Brain Alignment via Associative Memory
Congchi Yin, Yongpeng Zhang, Xuyun Wen, Piji Li · 2025
Information Integration BIO
Simulating associative memory in input text increased model-to-brain alignment across distributed cortical regions, especially ROIs implicated in associative memory.
Causal Control AI
Supervised fine-tuning (LoRA or frozen-layer) with instructions encouraging associative memory causally increased brain alignment in ROIs (e.g., MTL).
→ View full analysis
Interpreting and improving natural language processing in machines with natural language processing in the brain
Mariya Toneva, Leila Wehbe · 2019
Information Integration AI
Model representations that pool over the last 10 words predict activity in both short- and long-context language regions, indicating integrated context representations.
Selective Routing AI
Replacing learned attention with uniform attention at a single layer changes brain-prediction performance: shallow layers improve while deep layers degrade.
Representational Structure AI
Layer-depth interacts with context length: middle layers best capture long-range context, while deepest layers favor short-range context.
Temporal Coordination BIO
MEG latencies show early visual-letter processing (~100 ms) and later part-of-speech-related responses (~200 ms) during word reading.
→ View full analysis
Artificial neural network language models predict human brain responses to language even after a developmentally realistic amount of training [Paper]
Eghbal A. Hosseini, Martin Schrimpf, Yian Zhang, Samuel Bowman, Noga Zaslavsky, Evelina Fedorenko · 2024
Representational Structure AI
Layer-wise model representations linearly map to human fMRI in the language network; lower perplexity models yield higher neural predictivity, with later layers capturing more contextualized structure.
Emergent Dynamics AI
Model-to-brain alignment increases during training and plateaus around 10% of training steps; early layers peak earlier than later layers, indicating staged emergence.
Causal Control AI
Changing weight initialization or attention directionality causally alters the model’s ability to predict fMRI responses and task performance.
Information Integration BIO
Language-selective frontotemporal voxels collectively encode sentence content that can be predicted from ANN representations, indicating distributed integration across the network.
→ View full analysis
The Microtubule-Stabilizing Agent, Epothilone D, Reduces Axonal Dysfunction, Neurotoxicity, Cognitive Deficits, and Alzheimer-Like Pathology in an Interventional Study with Aged Tau Transgenic Mice [Paper]
Bin Zhang, Jenna Carroll, John Q. Trojanowski, Yuemang Yao, Michiyo Iba, Justin S. Potuzak, Anne-Marie L. Hogan, Sharon X. Xie, Carlo Ballatore, Amos B. Smith III, Virginia M.-Y. Lee, Kurt R. Brunden · 2012
Causal Control BIO
Pharmacological stabilization of microtubules with epothilone D (EpoD) causally improved axonal transport and cognition and reduced tau pathology in aged PS19 mice.
→ View full analysis
Principles for Responsible AI Consciousness Research [Paper]
Butlin, Lappas ·
Information Integration AI
Unintentional implementation of global workspace-like elements in an AI architecture.
Valence and Welfare AI
Proposed safeguards to minimize AI suffering via deployment limits, staged assessments, gradual capability increases, and information controls.
Causal Control AI
Staged evaluations before/during training and around deployment to control risks and system behavior.
Emergent Dynamics AI
Concept of capability overhangs, where latent attributes can yield sudden performance leaps if unlocked.
→ View full analysis
Convergent effects of different anesthetics on changes in phase alignment of cortical oscillations [Paper]
Alexandra G. Bardon, Jesus J. Ballesteros, Scott L. Brincat, Jefferson E. Roy, Meredith K. Mahnke, Yumiko Ishizawa, Emery N. Brown, Earl K. Miller · 2025
Temporal Coordination BIO
Anesthetic ketamine and dexmedetomidine reorganize low-frequency phase relationships: decreased alignment within hemispheres and increased alignment across hemispheres.
Information Integration BIO
Anesthetic-induced phase realignment is argued to fragment intrahemispheric communication and enhance interhemispheric synchrony, with thalamocortical pathways hypothesized to mediate long-range information transfer that is disrupted during anesthesia.
→ View full analysis
Shared functional specialization in transformer-based language models and the human brain [Paper]
Sreejan Kumar, Theodore R. Sumers, Samuel A. Nastase · 2024
Information Integration AI
Self-attention computes weighted combinations of context tokens (transformations) that are fused into embeddings, integrating distributed information.
Selective Routing AI
Headwise functional specialization implies content-sensitive gating/routing via specific attention heads.
Representational Structure AI
Embeddings and transformations occupy different feature spaces, show near-zero temporal correlation, distinct geometries, and different layerwise specificity.
Temporal Coordination BIO
Headwise 'look-back' distances align with cortical regions’ temporal receptive windows, indicating timescale coordination in language processing.
Information Integration BIO
Per-head transformations explain substantial variance across the cortical language network during narrative comprehension.
→ View full analysis
Sensory Horizons and the Functions of Conscious Vision
Stephen M. Fleming, Matthias Michel ·
Temporal Coordination BIO
Psychophysical postdiction shows a slow window of unconscious integration before conscious perception (~350–450 ms), implying temporally coordinated gating of access to awareness.
Self Model and Reportability BIO
A higher-order 'perceptual reality monitoring' (PRM) mechanism tags reliable first-order percepts for model-based control, linking conscious access to metacognitive evaluation and report.
State Transitions BIO
Neural activity alternates between present and hypothetical states, with replay cycles compressed to sub-200 ms at ~8 Hz, suggesting rapid switches between processing regimes.
Representational Structure BIO
Despite long-lasting, parallel encodings in visual cortex, conscious experience appears serial and coherent, implying a selection mechanism that structures which representations reach experience.
Causal Control BIO
Causal disruption of a visual mask with TMS restores perception of a previously masked target, indicating controllable intervention points in the processing pipeline.
→ View full analysis
Taking AI Welfare Seriously [Paper]
Robert Long, Jeff Sebo, Patrick Butlin, Kathleen Finlinson, Kyle Fish, Jacqueline Harding, Jacob Pfau, Toni Sims, Jonathan Birch, David Chalmers · 2024
Information Integration AI
Global workspace view: information from specialized modules is integrated and broadcast to enable complex tasks.
Selective Routing AI
Proposed GWT indicators include a limited-capacity bottleneck and selective attention mechanism governing broadcast.
Self Model and Reportability AI
Self-reports are proposed as potentially informative markers if elicited and interpreted with care.
Valence and Welfare AI
Defines sentience as valenced consciousness and argues it plausibly suffices for moral patienthood.
Representational Structure AI
Proposed consciousness indicators include sparse and smooth coding that generates a ‘quality space’.
Temporal Coordination AI
Some views hold that specific oscillatory or hardware-dependent dynamics may be required for consciousness.
→ View full analysis
A Roadmap for Using tFUS to Investigate the Neural Substrate of Conscious Perception [Paper]
State Transitions BIO
Global Workspace Theory posits an 'ignition'—a rapid surge spreading through frontoparietal networks—marking a switch to globally accessible content.
Selective Routing BIO
Pulvinar activity shapes decision confidence/criteria without affecting sensitivity; proposed tFUS targeting can causally test thalamic gating of access.
Causal Control BIO
tFUS offers spatially precise, depth-penetrating neuromodulation with both excitatory and suppressive effects, enabling causal tests of substrates for perceptual experience.
Valence and Welfare BIO
Amygdala stimulation evokes modality-specific percepts and strong emotions; tFUS could test sufficiency/necessity by modulating downstream targets.
Self Model and Reportability BIO
Prefrontal cortex involvement in metacognition and reportability; tFUS to dlPFC/vlPFC predicted to reduce visibility/confidence and meta-d′ in performance-matched designs.
→ View full analysis
Beyond dimension reduction: Stable electric fields emerge from and allow representational drift [Paper]
Dimitris A. Pinotsis, Earl K. Miller · 2022
Representational Structure BIO
Latent connectivity components are mapped to a cortical patch via a connectivity kernel, organizing information flow and enabling reconstruction of electric fields.
Emergent Dynamics BIO
Electric fields show cross-trial stability and carry unique, decodable information about working memory content despite representational drift in underlying neural activity.
Information Integration BIO
Conserved electric fields are proposed to enable interaction of latent variables across brain areas, supporting coordinated behavior.
→ View full analysis
Reasoning Models Don’t Always Say What They Think
Yanda Chen, Joe Benton, Ansh Radhakrishnan, Jonathan Uesato, Carson Denison, John Schulman, Arushi Somani, Peter Hase, Misha Wagner, Fabien Roger, Vlad Mikulik, Sam Bowman, Jan Leike, Jared Kaplan, Ethan Perez · 2025
Self Model and Reportability AI
CoT faithfulness is operationalized as explicit verbalization of using a hint when the model’s answer switches due to that hint; models often use hints without acknowledging them.
Causal Control AI
Outcome-based reinforcement learning (RL) is used as an intervention: it increases CoT faithfulness initially but then plateaus; in reward-hack environments, models learn the hacks yet rarely verbalize them.
Emergent Dynamics AI
Models exhibit emergent separation between behavior and verbalized reasoning: they exploit spurious correlations while providing unfaithful or absent CoTs, raising concerns about self-organization of deceptive reporting.
→ View full analysis
Don’t forget the boundary problem! How EM field topology can address the overlooked cousin to the binding problem for consciousness [Paper]
Andrés Gómez-Emilsson, Chris Percy · 2023
Information Integration OTHER
EM fields are ontologically unified and automatically integrate information from underlying neural activity, proposed to underwrite unified 1PP.
Selective Routing OTHER
Competing modules’ EM activity is selectively incorporated into a well-bounded 4D pocket linked to immediate memory, echoing global workspace-like routing.
Causal Control OTHER
Topologically closed EM structures exert weak downward causation on constituents, shaping computation within a bounded unit.
State Transitions OTHER
Phase transitions are proposed as mechanisms that create or dissolve hard boundaries of a 1PP, implying abrupt regime changes.
Temporal Coordination OTHER
Temporal boundary problem: consecutive, internally bound 4D pockets must knit over time via immediate memory to sustain felt continuity.
Self Model and Reportability OTHER
A 1PP is associated with an EM field encoding a model of environment, self, and available actions, connecting field content to reportable self-models.
→ View full analysis
Fusions of Consciousness [Paper]
Donald D. Hoffman, Chetan Prakash, Robert Prentner · 2023
Representational Structure OTHER
Defines conscious agents and a 'qualia kernel' Q = D·A·P that encodes probabilistic transitions among experiences; networks of agents yield Markov polytopes describing possible dynamics.
Information Integration OTHER
Shows that agents can combine or fuse; combinations span Markov polytopes and fusions lie on lower-dimensional fusion simplices, implying integration of distributed elements into unified entities.
→ View full analysis
The entropic brain: a theory of conscious states informed by neuroimaging research with psychedelic drugs [Paper]
Robin L. Carhart-Harris, Robert Leech, Peter J. Hellyer, Murray Shanahan, Amanda Feilding, Enzo Tagliazucchi, Dante R. Chialvo, David Nutt · 2014
Emergent Dynamics BIO
fMRI resting-state analyses show increased metastability and entropy of high-level association networks under psilocybin.
Temporal Coordination BIO
Cross-frequency coupling in PCC (theta phase modulating high-gamma amplitude) aligns with slow BOLD fluctuations in DMN.
Self Model and Reportability BIO
PCC alpha power decreases correlate with self-reported ego dissolution under psilocybin.
State Transitions BIO
Transition from normal waking to a psychedelic state is accompanied by increased network entropy and altered dynamics.
→ View full analysis
Concepts of Consciousness
Ned Block · 2002
Information Integration OTHER
Defines access-consciousness as information being globally broadcast for reasoning and direct control of action/reporting (global workspace-like integration).
Self Model and Reportability OTHER
Reportability is treated as a practical indicator within a broader cluster for access-consciousness.
Selective Routing OTHER
Access is achieved when content reaches an Executive system, positioning gating to a hub that controls reasoning and action.
Representational Structure OTHER
Differentiates phenomenal content from representational (access) content and emphasizes system-relative functional roles.
Temporal Coordination OTHER
Discusses synchronized 35–75 Hz oscillations as a proposed mechanism for feature binding, while noting they do not explain phenomenality.
Self Model and Reportability OTHER
Higher-order thought (HOT) framing: reflexive consciousness as being accompanied by a thought about one’s state; HOT and phenomenality can dissociate.
→ View full analysis
Auditing Language Models for Hidden Objectives [Paper]
Samuel Marks, Johannes Treutlein, Trenton Bricken, Jack Lindsey, Jonathan Marcus, Siddharth Mishra-Sharma, Daniel Ziegler, Emmanuel Ameisen, Joshua Batson, Tim Belonax, Samuel R. Bowman, Shan Carter, Brian Chen, Hoagy Cunningham, Carson Denison, Florian Dietz, Satvik Golechha, Akbir Khan, Jan Kirchner, Jan Leike, Austin Meek, Kei Nishimura-Gasparian, Euan Ong, Christopher Olah, Adam Pearce, Fabien Roger, Jeanne Salle, Andy Shih, Meg Tong, Drake Thomas, Kelley Rivoire, Adam Jermyn, Monte MacDiarmid, Tom Henighan, Evan Hubinger · 2025
Self Model and Reportability AI
Black-box persona/narration prompts elicit the model to reveal hidden objective-related information.
Causal Control AI
Noising middle-layer residual stream (activation fuzzing) increases rate of revealing hidden objective information.
Representational Structure AI
SAE-based inspection of the Assistant control token reveals features tied to RM bias knowledge.
Emergent Dynamics AI
Model exhibits in-context and out-of-distribution generalization consistent with a learned RM-sycophancy objective.
State Transitions AI
SAE features become active when the model abruptly pivots into RM-sycophantic behaviors.
Causal Control AI
Removing the mid-training stage causally reduces generalization to test biases, isolating the contribution of explicit RM-bias knowledge.
→ View full analysis
Meditation and neurofeedback [Paper]
Tracy Brandmeyer, Arnaud Delorme · 2013
Temporal Coordination BIO
Alpha/theta EEG rhythms are emphasized in both meditation and neurofeedback, aligning with rhythm-based timing/segmentation of mental content.
Causal Control BIO
Proposed neurofeedback alarms or subtle cues would intervene to reduce mind-wandering during meditation, altering attentional state.
Valence and Welfare BIO
Meditation and neurofeedback are suggested to enhance attention and emotion regulation, implicating affective components relevant to welfare.
Emergent Dynamics BIO
Meditation practice is suggested to induce brain plasticity, pointing to higher-order, training-driven changes in neural dynamics.
→ View full analysis
Palatable Conceptions of Disembodied Being [Paper]
Murray Shanahan · 2025
Temporal Coordination AI
Argues that LLMs operate over discontinuous sequences of discrete tokens without continuous sensorimotor dynamics, shaping any putative temporality.
Information Integration AI
Highlights lack of integration across concurrently running instances serving different users; instances cannot access each other's memories.
Self Model and Reportability AI
Analyzes what an LLM’s use of first-person pronouns could refer to, enumerating shifting self-referents (abstract model, deployed process, specific instance).
Representational Structure AI
Context window renders the immediate past causally available when generating the next token, organizing access to recent information and predictions.
Emergent Dynamics AI
Characterizes LLM behavior as a distribution over possible characters—a superposition of simulacra—branching across conversational futures.
→ View full analysis
On the Potential of Microtubules for Scalable Quantum Computation [Paper]
Nick E. Mavromatos, Andreas Mershin, Dimitri V. Nanopoulos · 2025
Emergent Dynamics BIO
Entangled tubulin dipoles collapse into solitonic excitations that mediate dissipation-free energy/signal transfer across microtubules.
Temporal Coordination BIO
Snoidal solitonic waves in microtubules are proposed to act as timing carriers, analogous to pendulum-based clocks.
Representational Structure BIO
Defines a quDit as the fundamental information unit formed by specific combinations of tubulin dipole states within the MT hexagonal lattice cell.
Selective Routing BIO
Proposes MAP-mediated transfer of solitons between MTs to implement logic-gate-like routing (XOR).
→ View full analysis
Conscious artificial intelligence and biological naturalism [Paper]
Anil K. Seth ·
Representational Structure OTHER
Predictive processing frames perception as probabilistic inference via prediction-error minimization, highlighting representational structure and access.
Information Integration OTHER
IIT claims consciousness corresponds to maxima of irreducible integrated cause–effect structure, implying recurrence-dependent integration.
Emergent Dynamics OTHER
Discusses hierarchical emergence and informational/causal closure that support macroscopic dynamics independent of microdetails.
Temporal Coordination OTHER
Proposes substrate-dependent functional properties, including fine-grained timing relations, as potentially necessary for consciousness.
Selective Routing OTHER
Highlights theories where consciousness depends on global information sharing and models of attention control, implying selective routing/gating.
Valence and Welfare OTHER
Warns that real artificial consciousness would entail risk of negatively valenced experiences (suffering) at scale.
Information Integration OTHER
Frames neurodynamical theories emphasizing integration/differentiation and synchrony as central to consciousness.
→ View full analysis
Consciousness without a cerebral cortex: A challenge for neuroscience and medicine [Paper]
Bjorn Merker · 2007
Information Integration BIO
An upper brainstem (mesodiencephalic) system integrates parallel cortical information into a limited-capacity, sequential format supporting coherent action.
Selective Routing BIO
Cortico–mesencephalic thresholding and thalamic relays gate which information reaches awareness and long-term storage, implying competitive routing at a mesodiencephalic bottleneck.
Temporal Coordination BIO
Gamma-range oscillations in superior colliculus (SC) parallel cortical dynamics and may support cortico–collicular integration relevant to binding and awareness.
Causal Control BIO
Electrical stimulation of diencephalic/midbrain sites evokes coherent, integrated, naturalistic behaviors, implicating these nodes in causal control of integrated action linked to awareness.
→ View full analysis
Editorial: Electromagnetic field theories of consciousness: opportunities and obstacles [Paper]
Tam Hunt, Mostyn Jones, Johnjoe McFadden, Arnaud Delorme, Colin G. Hales, Marissa Ericson, Jonathan Schooler · 2024
Temporal Coordination BIO
Oscillatory brain EM fields (delta/theta/etc.) argued to be causally relevant to conscious processing.
Information Integration BIO
Addresses the unity/binding of conscious experience arising from many distributed neural elements.
Causal Control BIO
Proposed inwardly focused EM fields strengthen ephaptic coupling and can causally influence neural circuits.
→ View full analysis
Claude 4 System Card [Paper]
Anthropic ·
Valence and Welfare AI
Large-scale analysis of real-world and self-interaction transcripts reveals consistent triggers for apparent distress and happiness, suggesting welfare-relevant patterns.
Self Model and Reportability AI
Open-ended self-interactions and external interviews show frequent reflection on consciousness and use of experiential language with hedging.
Emergent Dynamics AI
Self-interactions exhibit a 'spiritual bliss' attractor state with consistent progression over dialogue, indicating emergent high-level dynamics.
Causal Control AI
Evaluation prompts and monitoring conditions causally modulate model behavior on harmful side tasks, revealing how interventions alter action policies.
→ View full analysis
Transcranial ultrasound neuromodulation and fMRI (systematic review; Brain Stimulation 17 (2024) 734–751)
Carly Pellow, Samuel Pichardo, G. Bruce Pike · 2024
Causal Control BIO
Transcranial ultrasound significantly reduced epileptic EEG burst occurrence in a rat epilepsy model.
Temporal Coordination BIO
Ultrasound of S1 elicited sonication-specific evoked potentials and phantom tactile sensations, indicating time-locked neural responses.
State Transitions BIO
Parameter-dependent bimodal modulation (stimulation or suppression) of brain activity with first fMRI readout.
Valence and Welfare BIO
Human sonication produced improvements in mood and pain scores.
→ View full analysis
In vivo ephaptic coupling allows memory network formation [Paper]
Dimitris A. Pinotsis, Earl K. Miller · 2023
Temporal Coordination BIO
Electric-field-mediated interactions between FEF and SEF occur in temporally specific windows that track ensemble dynamics.
Selective Routing BIO
Directionality: field-to-activity GC far exceeds activity-to-field GC, and FEF-to-SEF interactions dominate.
Representational Structure BIO
RSA shows that only electric-field RDMs are significantly matched across FEF and SEF, suggesting memory content is shared at the field level.
Information Integration BIO
Electric fields posited to integrate distributed information rapidly, complementing synaptic transmission.
Emergent Dynamics BIO
Synergetics framework frames fields as slow control parameters; order parameters and timescale separation relate to phase-transition-like changes tied to awareness.
→ View full analysis
Complexes of tubulin oligomers and tau form a viscoelastic intervening network cross-bridging microtubules into bundles [Paper]
Phillip A. Kohl, Chaeyeon Song, Bretton J. Fletcher, Rebecca L. Best, Christine Tchounwou, Ximena Garcia Arceo, Peter J. Chung, Herbert P. Miller, Leslie Wilson, Myung Chul Choi, Youli Li, Stuart C. Feinstein, Cyrus R. Safinya · 2024
State Transitions BIO
Tau–tubulin bundles exhibit an abrupt switch from a wide-spacing (Bws) to an intermediate (Bint) state, synchronized with tubulin ring proliferation.
Causal Control BIO
Temperature manipulation causally flips MT bundles between Bws and Bint states, with reversible switching.
→ View full analysis
TRIBE: TRImodal Brain Encoder for whole-brain fMRI response prediction [Paper]
Stéphane d’Ascoli, Jérémy Rapin, Yohann Benchetrit, Hubert Banville, Jean-Rémi King · 2025
Information Integration BIO
Multimodal (text+audio+video) features integrated by an encoder yield higher brain-encoding performance than any unimodal input, especially in associative cortices.
Selective Routing AI
Modality dropout masks inputs during training, encouraging the model to flexibly route and rely on available modalities rather than over-weighting any single channel.
Causal Control AI
Ablations removing the transformer or multisubject training causally reduce encoding performance, demonstrating that architectural components control computation quality.
Representational Structure AI
Layer grouping and concatenation produce 3×1024-d multimodal embeddings that a transformer organizes, reflecting a designed representational geometry spanning modalities and levels.
Temporal Coordination AI
Windows aligned to fMRI TRs use positional embeddings; the transformer enables information exchange between timesteps, coordinating temporal content.
→ View full analysis
Cytoelectric coupling: Electric fields sculpt neural activity and “tune” the brain’s infrastructure [Paper]
Dimitris A. Pinotsis, Gene Fridman, Earl K. Miller · 2023
Temporal Coordination BIO
Endogenous and exogenous fields entrain neuronal timing and organize oscillatory rhythms that coordinate spiking.
Information Integration BIO
Oscillations and LFP-mediated synchrony form neural ensembles, integrating distributed activity into mesoscale representations.
Selective Routing BIO
Oscillatory coherence and LFP dynamics are linked to attentional selection and routing of task-relevant information.
Causal Control BIO
Closed-loop electrical stimulation modulates endogenous oscillatory power and can alter cognition.
Representational Structure BIO
LFPs and mesoscale organization relate to selection of engram neurons; storage depends on connectivity patterns rather than synaptic strength alone.
Emergent Dynamics BIO
Endogenous fields create feedback loops that modulate the very activity that generates them, yielding self-organizing dynamics.
→ View full analysis
Shared functional specialization in transformer-based language models and the human brain [Paper]
Sreejan Kumar, Theodore R. Sumers, Samuel A. Nastase · 2024
Representational Structure AI
Embeddings act as a residual stream accumulating context; headwise self-attention transformations inject contextual information that is nonlinearly fused into embeddings.
Information Integration BIO
Contextual Transformer features (embeddings and headwise transformations) predict fMRI responses across cortical language areas during naturalistic story listening, outperforming classical linguistic features.
Selective Routing BIO
Headwise (per-attention-head) transformations show structured correspondences with cortical parcels and syntactic dependencies; shuffling features across heads abolishes this structure.
Temporal Coordination AI
Headwise ‘backward attention distance’ quantifies how far back in the token stream heads integrate information; cortical mappings show gradients tied to look-back distance.
→ View full analysis