Does an Agent's History Tell You When Compaction Will Hurt? A Modest, Bounded Effect on the TRACE Paired-Replay Corpus

Key points

• Many long-horizon agents compact their context on a global rule, usually a token budget, blind to what the agent was doing.1
• We ask whether the agent's recent behaviour predicts when a compaction will hurt.1
• TRACE's public corpus of 590 harness-triggered AppWorld compaction boundaries replays each boundary from a re-executed prefix state under the pre-compaction context and under the summary, and records the burden of the next actions: calls that error or repeat a call already made.1
• Whether the best trigger beats a token-budget rule at matched retention cannot be evaluated on the release.1

Sources (1)

• [1]Does an Agent's History Tell You When Compaction Will Hurt? A Modest, Bounded Effect on the TRACE Paired-Replay Corpus
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 6, 05:26 PM

  We find that pre-boundary history predicts post-compaction harm only weakly.

  Many long-horizon agents compact their context on a global rule, usually a token budget, blind to what the agent was doing.

Extractive summary: sentences quoted from the sources.