Safer Reasoning Traces: Measuring and Mitigating Chain-of-Thought Leakage in LLMs

source_context

{
  "surface": "signal_canvas",
  "mode": "paper",
  "query": "Safer Reasoning Traces: Measuring and Mitigating Chain-of-Thought Leakage in LLMs",
  "normalized_query": "2603.05618",
  "route": "/signal-canvas/safer-reasoning-traces-measuring-and-mitigating-chain-of-thought-leakage-in-llms",
  "paper_ref": "safer-reasoning-traces-measuring-and-mitigating-chain-of-thought-leakage-in-llms",
  "topic_slug": null,
  "benchmark_ref": null,
  "dataset_ref": null
}