Line of inquiry
Inquiring lines›Where does language-model reasonin…›How do language models represent m…›this line of inquiry
Do language models develop causal world models or rely on statistical patterns?
A broader line of inquiry — a family of 27 specific questions the research asks around this. Follow one into its inquiring-line page, or move sideways to a related line below.
Questions in this line of inquiry 27
Specific inquiring lines the field asks around this — ordered from the most general framing down to the most specific angle.
- Can language models develop world models that ground meaning in causal reality?
- Do language models build world models or just task-specific heuristics?
- Can language models learn internal world models without explicit environment specifications?
- Do language models need words to think or just latent structure?
- Can LLM semantic representations exist without causally influencing their generation output?
- Do LLMs rely on surface statistical patterns instead of causal structure?
- What prevents LLM representations from causally influencing generation outputs?
- Why do language models capture individual differences in cognitive behavior?
- Do LLMs need world models to make accurate predictions?
- Can external actions provide causal necessity that language models lack?
- Do LLMs genuinely internalize human psychological structure or match surface patterns?
- How do internal representations compare to human cognitive structures?
- Can language models generate plausible latent thoughts without human annotation?
- Does directional knowledge failure indicate shallow pattern matching over deep representation?
- What data presentation structures enable LLMs to learn decision-making from examples?
- Can frozen world models from training cutoff remain adequate for real-world reasoning?
- How can we probe LLM representations in channels that training did not target?
- How should we rethink the symbolism versus connectionism debate in light of LLMs?
- How do world models create indirect causal grounding without physical environment contact?
- Can external summarization solve exploration problems in complex real-world environments?
- What empirical evidence supports the Learning Law on real language models?
- Can models track dynamic mental state changes better than static beliefs?
- Why do language models reproduce human EPA structure despite different architecture?
- Why must world models be nested rather than flat and uniform?
- How do world models decompose between representation of facts versus generative mechanisms?
- Why does integrating world models with decision-making systems matter?
- Does next-state prediction alone build mechanistic world models or just sophisticated interpolation?