Signals and Queries: Token-Efficient Agentic Context Discovery

A support or data assistant repeats work across related customer requests.

What you would change

Try the paper's memory change in a replay of past tasks, with separate memory per customer.

Try the paper's memory change in a replay of past tasks, with separate memory per customer.

Compare against

Your current memory pipeline plus a no-memory control using the same agent.

1. Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.
2. Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.
3. Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.
4. Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test.

What to measure

Task success, stale memories, retrieval misses, response latency, and all memory maintenance costs.

When to keep it

Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation.

recommendation

adopt

research area

memory and context

technical change

Per-policy embedding classifier as pre-filter for agent policies

published

2026-10-21

The method requires only embedding APIs and a logistic classifier, fits a 24 GB GPU or API-only setup with no per-turn LLM cost, and directly reduces production token spend, so it is adoptable now provided recall is validated on the reader's own policy set.

1. claim 01Paperraft inference
   Read the brief interpretation

   S&Q replaces per-turn LLM-based relevance evaluation of every behavioral policy with an offline-trained, per-policy logistic classifier over embeddings that filters candidates before any LLM call. The cost is an offline generation step (activating examples and hard negatives per policy), classifier training and maintenance per policy, and a small risk of false negatives that downstream stages never see. It can fail when user behavior drifts from the offline training examples, when new policies lack sufficient generated examples, or when embedding quality is poor for the domain, since a wrongly discarded policy is silently excluded from that turn.

   This is a Paperraft inference. It is not pinned to a verified PDF passage.

   Inspect the source
2. claim 02Paperraft inference

   Removes at least 30% of candidate policies before downstream LLM evaluation while preserving high recall, with higher precision than top-K and cosine-threshold retrieval at every evaluated recall level.

   This is a Paperraft inference. It is not pinned to a verified PDF passage.

   Inspect the source