Tag
This paper presents a two-stage LLM pipeline for extracting and triangulating causal evidence from humanitarian crisis reports, achieving strong F1 scores on a ReliefWeb dataset and proposing a Level-of-Evidence score for cross-context convergence.
This paper tests the assumption that attention weights reveal what a model actually depends on for its output, finding that attention and causal dependence often disagree. They propose using causal evidence sets obtained via intervention masking as supervision for sparse attention routers, achieving near-perfect accuracy on retrieval tasks where attention-distilled routers fail.
This paper provides causal evidence that large language models acquire negative linguistic knowledge (what not to say) through statistical preemption, a mechanism from Construction Grammar, by showing that manipulating competing-form frequencies via fine-tuning shifts preemption behavior in predicted directions.