Model Releases
Self-Compacting Language Model Agents
Long agent traces composed of chains of thought and tool calls accumulate stale content that anchor subsequent generations, and eventually outgrow the context window. Existing scaffolds mitigate it wi
Long agent traces composed of chains of thought and tool calls accumulate stale content that anchor subsequent generations, and eventually outgrow the context window. Existing scaffolds mitigate it with fixed-interval compaction triggered at a token threshold. Such triggers pay no heed to trajectory structure, risking discard of partial results mid-derivation or mid-search. We propose SELFCOMPACT, a scaffold that allows the model itself to decide when and how to compact. Specifically, it pairs two inference-time elements: (i) a compaction tool the model invokes to summarize the accumulated…
Related
- Addressable Recall Compaction for Long Context-Window Control in AI Agents
- Parallel Context Compaction for Long-Horizon LLM Agent Serving
- Learning What Not to Forget: Long-Horizon Agent Memory from a Few Kilobytes of Learning
- Context as an Environment: Programmatic Context Management for Long-Horizon Agents
- LiveMem: Maintaining Memory State Continuity in Long-Running LLM Inference
Source: Apple ML Research | 2026-10-12