SYNTHESIS NOTE
Topics›Context Engineering›this note

Can an external manager handle context for frozen agents?

Exploring whether a separate trained system can effectively manage a frozen agent's context window. This matters because many deployed agents are closed-source and can't be retrained, yet they suffer from context degradation.

Synthesis note · 2026-06-03 · sourced from Context Engineering
How does test-time scaling work for individual research agents?

Long-horizon agents accumulate context — tool results, intermediate reasoning — until stale content obscures salient evidence, amplifies positional bias, and degrades decisions. Prior fixes put the burden of managing context on the agent itself (agent-side control, or fixed summarization), which requires training the agent and is impractical for closed-source agents, and ignores that different agents need different strategies.

AdaCoM separates the concern entirely: train an external LLM to manage the context of a frozen agent through flexible modification actions and end-to-end RL. The manager prunes stale content while preserving task constraints and progress, improving diverse agents on web-search and deep-research benchmarks and transferring to unseen agents of similar capability.

The most useful finding is a fidelity–reliability trade-off. Agents with higher vanilla ReAct performance benefit from higher-fidelity context preservation — they can use more detail well. Lower-performing agents require more aggressive compression to stay within a reliable reasoning regime. The right amount of context is not a property of the task alone; it is indexed to the agent's own competence. This means context management is not one universal policy but a per-agent calibration — consistent with Does fixed sparsity work for all sequence lengths?, where the optimal budget is also conditional rather than fixed.

Inquiring lines that read this note 49

This note is a source for these research framings, grouped by the broader line of inquiry each explores. Scan the bold lines of inquiry; follow any specific question forward.

How do users build accurate mental models when agent reasoning is partially hidden? Why do AI agents struggle to generalize and improve reliably? How should agents manage memory consolidation and reuse over time? How does compressing or distilling information affect model learning and generalization? How should architects optimize KV cache memory and compute tradeoffs for inference? Can external memory modules outperform model weight consolidation for long-term retention? What distinguishes genuine linguistic understanding from mere pattern matching? Do evolved harness optimizations generalize beyond their source benchmarks? What makes decomposed prompts more reliable than monolithic instructions? Why does consolidating agent memories often degrade performance? How should persistent agent-authored code be governed and shared? How should control and autonomy be allocated in human-AI workplace collaboration? How should retrieval systems adapt to different reasoning problems? How do attention patterns affect model performance across varying context lengths? Do harness improvements deliver capability gains comparable to model scaling? Do fine-tuning methods modify core model capabilities or just outputs? What causes models to fail in specialized domains versus generally? How should agent identity and authorization work across organizational boundaries? Why do autonomous agents report success despite actually failing? What organizational factors determine whether firms successfully adopt and integrate AI?

Related concepts in this collection 4

This note in its neighbourhood — explore the map, then jump to a related concept in the list below.

Concept map
17 direct connections · 135 in 2-hop network ·medium cluster Open in graph ↗

Click a node to walk · click center to open · click Open in graph to see this note in the full knowledge graph

your link semantically near linked from elsewhere

Related papers in this collection 8

Papers most semantically related to this note, ranked by cosine similarity in the embedding space.

Original note title

context management can be offloaded to a trained external manager for a frozen agent and optimal compression depends on the agent's own reliability