salitrap-7a65acd1·1 events·first seen Aliases: SaliTrap
Researchers introduce 'Salience Bias', a failure mode where LLMs over-prioritize explicit numerical or surface-level distractors in prompts, causing them to ignore implicit commonsense prerequisites. They construct SaliTrap, a benchmark across four distractor dimensions, and evaluate 12 state-of-the-art LLMs, finding all suffer significantly from this bias. Crucially, context-free knowledge probing recovers over 90% of failures, demonstrating the issue is knowledge suppression rather than knowledge absence. Lightweight inference-time prompting substantially closes the gap without retraining, relocating the bottleneck from model competence to elicitation.