arXiv cs.AI
  • Score 83
  • Official

Agents Cannot Hand Over What They Have Not Yet Understood

arXiv:2611.18642v1 Announce Type: new Abstract: Agents that work beyond a single context window, or in teams, hand their work to a successor as a short written summary. Evaluation has focused on whether an agent can solve a task, and relatively little is known about how much of what an agent has learned survives being written down. We study this in 56 interactive worlds with hidden rules, where an

So what

What this event means by reading role—not a longer recap.

  • BuilderIf you chain agents via short summaries, write each note from the full history rather than the previous note, which recovers 70-82% of the handoff loss.
  • ResearcherThis quantifies a handoff loss across six models and decomposes it into agent change, single compression, and repeated compression components.

Score dimensions

Higher total means read first. Each bar is one factor we use to rank the system pool. How we score

RelevanceHow tightly this is about AI.
82
ImpactHow much this could change the field or the market.
55
NoveltyHow new this is versus a recap.
70
CredibilityHow much we trust the source.
92
ActionabilityWhether a reader can do something with it.
45