REVIEW 14 cited by
The Consciousness Prior
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
The Consciousness Prior
read the original abstract
A new prior is proposed for learning representations of high-level concepts of the kind we manipulate with language. This prior can be combined with other priors in order to help disentangling abstract factors from each other. It is inspired by cognitive neuroscience theories of consciousness, seen as a bottleneck through which just a few elements, after having been selected by attention from a broader pool, are then broadcast and condition further processing, both in perception and decision-making. The set of recently selected elements one becomes aware of is seen as forming a low-dimensional conscious state. This conscious state is combining the few concepts constituting a conscious thought, i.e., what one is immediately conscious of at a particular moment. We claim that this architectural and information-processing constraint corresponds to assumptions about the joint distribution between high-level concepts. To the extent that these assumptions are generally true (and the form of natural language seems consistent with them), they can form a useful prior for representation learning. A low-dimensional thought or conscious state is analogous to a sentence: it involves only a few variables and yet can make a statement with very high probability of being true. This is consistent with a joint distribution (over high-level concepts) which has the form of a sparse factor graph, i.e., where the dependencies captured by each factor of the factor graph involve only very few variables while creating a strong dip in the overall energy function. The consciousness prior also makes it natural to map conscious states to natural language utterances or to express classical AI knowledge in a form similar to facts and rules, albeit capturing uncertainty as well as efficient search mechanisms implemented by attention mechanisms.
Forward citations
Cited by 14 Pith papers
-
Hidden APIs in Language Models: Discovering Reusable Causal Interfaces from Forked Futures
A Shared hidden-state interface beats Local, Mixture, and Distributed alternatives under held-out causal description length in Qwen2.5-1.5B and Llama-3-8B, with transplantation and mediation evidence of reuse.
-
Verbalizable Representations Form a Global Workspace in Language Models
Language models represent their current reasoning in a small, readable set of verbalizable vectors (the J-space) that functions like a global workspace.
-
Bayesian updates from coalgebraic determinisation
Unifilarisation of stochastic Mealy machines is an instance of coalgebraic determinisation over monads with support structure, producing causal stochastic behaviours rather than Moore-style output distributions.
-
EasyLens: A Training-Free Plug-and-Play Subtle-Lesion Representation Amplifier for Medical Vision-Language Models
Prior-minimal multi-agent RL agents develop indexical encoding, persistent self-state, and an echo-mismatch self-monitoring circuit that vanishes when the echo affordance is removed during training.
-
ViperGPT: Visual Inference via Python Execution for Reasoning
ViperGPT generates executable Python code to compose pre-trained vision-and-language modules into programs that answer visual queries, reaching state-of-the-art results with no additional training.
-
EasyLens: A Training-Free Plug-and-Play Subtle-Lesion Representation Amplifier for Medical Vision-Language Models
EasyLens introduces a plug-and-play amplifier that uses pathology-anatomy prototypes and morphology-guided residual enhancement to boost subtle-lesion cues in frozen medical VLMs.
-
Generative Recursive Reasoning
GRAM is a latent-variable generative model that performs recursive reasoning via stochastic trajectories, trained with amortized variational inference to support multi-hypothesis reasoning and unconditional generation.
-
Generative Recursive Reasoning
GRAM turns recursive latent reasoning into a generative probabilistic model via stochastic trajectories and amortized variational inference, claiming better performance on structured reasoning tasks than deterministic...
-
When to Think Fast and Slow? AMOR: Adaptive Entropy Gate for Hybrid Models
AMOR uses output entropy to gate attention in recurrent hybrids, matching full attention performance at roughly 22% attention invocations across 180M-1.5B models.
-
Dynamical Mechanisms for Coordinating Long-term Working Memory Based on the Precision of Spike-timing in Cortical Neurons
Cortical traveling waves may use precise spike timing and temporary STDP to form a second network that sustains long-term working memory for hours.
-
Thoughtseeds as Latent Causes: A Dual-Process Computational Phenomenology of Focused-Attention Meditation
A three-layer active-inference simulation of focused-attention meditation reproduces expert–novice differences, yet the key differences are pre-specified in hand-set priors rather than derived.
-
Emergent Language as an Approach to Conscious AI
Agents in a minimal multi-agent RL setup develop self-referential communication and an echo-mismatch detection circuit that emerges from environmental affordances rather than task structure or architecture.
-
Dynamical Mechanisms for Coordinating Long-term Working Memory Based on the Precision of Spike-timing in Cortical Neurons
Coordinated millisecond-precision spike timing and traveling waves may enable a second tier of cortical activity for long-term working memory via sustained STDP.
-
AI Consciousness and Existential Risk
Consciousness does not directly predict AI existential risk unlike intelligence, though it may indirectly affect risk through alignment or capability requirements.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.