Line of inquiry
Inquiring lines›How do language models learn and r…›How do language models learn and w…›this line of inquiry
Can language models reason beyond surface pattern matching?
A broader line of inquiry — a family of 91 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 91
Specific inquiring lines the field asks around this — ordered from the most general framing down to the most specific angle.
- Can language models reason without relying on learned semantic patterns?
- Why do language models imitate reasoning form without abstract inference capability?
- Do LLMs learn surface patterns instead of genuine linguistic structure?
- Why do LLMs fail at semantic generalization despite grammatical accuracy?
- Can language models perform genuine symbolic reasoning without semantic grounding?
- Can LLMs reason through semantics without understanding causal mechanisms?
- Can LLM semantic representations exist without causally influencing their generation output?
- Can language models reason without relying on surface level pattern matching?
- Do language models build world models or just task-specific heuristics?
- Do language models learn surface patterns that appear generalizable but actually fail under shift?
- Can language models learn internal world models without explicit environment specifications?
- Do language models need words to think or just latent structure?
- Can explicit connectives compensate for missing intentional tracking in LLMs?
- Do language models exhibit the same causal biases that humans show?
- Why do language models fail at implicit discourse relations while handling explicit connectives?
- Why do language models struggle with context-dependent pragmatic interpretation?
- Why do LLMs perform better on explicit discourse connectives than implicit relations?
- Do language models actually learn linguistic structure or just surface statistics?
- How does implicit meaning processing limit LLM pragmatic reasoning?
- How faithful are natural language explanations from LLMs really?
- Do language models encode deep syntactic structure or only surface-level patterns?
- Can language models translate theorems faithfully without semantic loss?
- Why do LLMs struggle to translate natural language into logical formalizations?
- Can language models develop world models that ground meaning in causal reality?
- Can LLMs translate between natural language and formal logic faithfully?
- Why do explicit discourse connectives help LLMs but implicit relations cause failures?
- Why do language models capture individual differences in cognitive behavior?
- Can LLMs decode their own hidden activations into natural language?
- Can language models perform purely symbolic reasoning when semantics are removed?
- Do LLMs compute scalar implicature differently across conversational contexts?
- Can language models keep secrets and control information strategically?
- Why do language models fail at planning despite understanding strategies?
- How does semantic grounding differ between human minds and language models?
- How does structural depth in sentences predict LLM annotation accuracy?
- Why do LLMs fail at faithful autoformalisation of reasoning problems?
- Do LLMs rely on surface heuristics instead of learning recursive grammar rules?
- Why do language models fail when semantic content is stripped away?
- What communicative optimization principles do language models fail to acquire?
- Can LLMs infer situational context the way humans do pragmatically?
- How do fixed pragmatic templates prevent models from understanding context?
- Do LLMs learn linguistic generalizations or just surface-level frequency patterns?
- Why do large language models still have systematic blind spots with complex structures?
- Can LLMs infer implicit meaning without surface linguistic markers?
- What semantic information is necessary to preserve for sound LLM reasoning?
- Can language models execute iterative numerical methods in latent space?
- Can language acquisition analogies mislead us about how models actually learn?
- Can language models distinguish explicit from implicit discourse relations?
- Can language models generate plausible latent thoughts without human annotation?
- Why do LLMs fail at implicit elements in literary and poetic text?
- Do metaphors work by decoupling meaning from linguistic associations?
- Does chain-of-thought prompting overcome implicit meaning deficits in text analysis?
- Do LLMs learn abstract grammar or culturally situated discourse patterns instead?
- Why do LLMs achieve only 24 percent accuracy on implicit discourse relations?
- Why do LLMs produce semantically acceptable but pragmatically disengaged responses?
- Why do explicit discourse connectives work when implicit relations fail?
- How do internal representations compare to human cognitive structures?
- Can LLMs improve at metaphor if they handle decoupled semantics better?
- Why does LLM compression eliminate causal grounding in conceptual representations?
- What causes language models' strategic rationality to decline with increased game complexity?
- Why do language models treat presupposition triggers as categorical patterns?
- Why do LLMs generate logical forms without preserving semantic content?
- Why do LLMs fail to actively reject false presuppositions in conversation?
- How do LLMs handle false presuppositions embedded in user questions?
- Can language models learn to diversify their discourse-level narrative patterns over time?
- Can we use LLM language without adopting LLM assumptions?
- How does the symbol grounding problem apply to artificial language systems?
- Can presupposition projection strength vary by context in embeddings?
- Why do LLMs miss new scientific ideas before they enter formal literature?
- Can language models understand the implicit emotional intent behind questions?
- How does syntactic encoding relate to semantic feature representation?
- What specific linguistic features cause LLMs to fail at trivial entailment?
- How do humans learn language through communication differently than LLM text prediction?
- How do LLMs lose information when translating natural language to formal logic?
- Why are false presuppositions harder to spot when they sound plausible?
- How does context collapse affect what language models can meaningfully communicate?
- How do embedding contexts like presupposition triggers affect LLM entailment reasoning?
- What empirical evidence supports the Learning Law on real language models?
- Why does hypothesis attestation bias exist separately from frequency bias in NLI?
- Why do language models reproduce human EPA structure despite different architecture?
- Can complexity-stratified testing reveal whether LLMs understand grammatical structure?
- Can language models ground clarifications without vision and kinesthetic modalities?
- What other latent LLM capabilities remain inactive without explicit activation cuing?
- Can LLMs identify implicit metaphoric mappings that require pragmatic inference?
- Can LLMs compute how presuppositions project through embedded clauses?
- Does DPO training with coreference chains teach spontaneous convention formation?
- What is the difference between learning discourse patterns and learning abstract language?
- How do LLMs translate informal prose into logically correct formal specifications?
- Can language models adapt irony detection to specific communicative contexts?
- How does bidirectional entailment distinguish semantic equivalence from token similarity?
- Can a system without an addressee ever truly tell a joke?
- Why do LLMs choose surface-order quantifier scope over contextually correct readings?