Larger LLMs acquire basic situation modeling before mentalizing on false-belief tasks, with performance depending on size, training volume, and post-training, yet remaining sensitive to non-factive verbs and agent knowledge states.
Trends in Cognitive Sciences , publisher =
5 Pith papers cite this work, alongside 8 external citations. Polarity classification is still indexing.
5
Pith papers citing it
8
external citations · OpenAlex
years
2026 5verdicts
UNVERDICTED 5representative citing papers
Intrinsic computational functionalism uses system-intrinsic instantiation (C1) and causal-dynamical organisation under intervention (C2) to identify observer-independent computational structures for consciousness via a three-tier decomposition of identification work.
Homeostatic coupling, not mere partner observation, produces food-sharing in a self-directed recurrent ALife agent with no partner-welfare reward.
Type-A Biological Naturalism is untestable because it decouples consciousness from behavior; Type-B is testable and compatible with computational functionalism, but both require linking consciousness to information processing.
Direct research on AI consciousness is intractable, so the field should prioritize studying perceived AI consciousness and its societal consequences.
citing papers explorer
-
Developmental Trajectories of Situation Modeling and Mentalizing in Transformer Language Models
Larger LLMs acquire basic situation modeling before mentalizing on false-belief tasks, with performance depending on size, training volume, and post-training, yet remaining sensitive to non-factive verbs and agent knowledge states.
-
Intrinsic Computational Functionalism: From Observer-Relative Maps to Observer-Independent Structures
Intrinsic computational functionalism uses system-intrinsic instantiation (C1) and causal-dynamical organisation under intervention (C2) to identify observer-independent computational structures for consciousness via a three-tier decomposition of identification work.
-
Prosociality by Coupling, Not Mere Observation: Homeostatic Sharing in an Inspectable Recurrent Artificial Life Agent
Homeostatic coupling, not mere partner observation, produces food-sharing in a self-directed recurrent ALife agent with no partner-welfare reward.
-
What biology can, and cannot, tell us about conscious AI
Type-A Biological Naturalism is untestable because it decouples consciousness from behavior; Type-B is testable and compatible with computational functionalism, but both require linking consciousness to information processing.
-
AI and Consciousness: Shifting Focus Towards Tractable Questions
Direct research on AI consciousness is intractable, so the field should prioritize studying perceived AI consciousness and its societal consequences.