arXiv AI September 24, 2026 By Mingxuan Wang, Bo Wang, Fei Luo, Guorun Yao, Chao Ning, Yinglong Guo, Hongyue Chen, Yanbiao Ma, Jungong Han

DRSR: Learning Set-Level Deletion Risk for Efficient Long-Horizon Agents

Read the original on arXiv AI →

The paper introduces Direct Relational Set‑Risk Pruning (DRSR), a method for compressing the history of long‑horizon language‑model agents by selecting deletion sets based on risk constraints rather than independent unit scores. DRSR builds counterfactual supervision offline, then uses a lightweight scorer to predict set‑level harm during deployment, removing the largest safe set while respecting recency, protocol, and budget limits. Experiments on WorkBuddyBench Full260 and Eval40 show that DRSR improves mean reward from 0.699 to 0.802 and reduces token usage by over 20%, with further analyses highlighting the importance of decision‑conditioned relations, retained context, pair interactions, and abstention.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Hugging Face Trending Papers

Sep 23

DRSR: Learning Set-Level Deletion Risk for Efficient Long-Horizon Agents

Long-horizon language-model agents accumulate reasoning traces, tool exchanges, and observations whose relevance changes with the current decision. Existing compression strategies often score historic...

agents efficiency safety

arXiv Machine Learning

Jun 5

SHRED: Retain-Set-Free Unlearning via Self-Distillation with Logit Demotion

arXiv:2605. 07482v2 Announce Type: replace Abstract: Machine unlearning for large language models (LLMs) aims to selectively remove memorized content such as private data, copyrighted text, or hazardous knowledge, without costly full retraining.

By Zizhao Hu, Ameya Godbole, Johnny Tian-Zheng Wei, Mohammad Rostami, Jesse Thomason, Robin Jia

llms efficiency benchmarks

arXiv AI

Sep 1

TRACER: Per-Tool Context Retention for LLM Agents via Consequence-Attributed Reinforcement Learning

arXiv:2608.29363v1 Announce Type: new Abstract: Enterprise data agents answer business queries by chaining many tool calls over multiple reasoning steps, routinely accumulating hundreds of thousands...

By Ziqi Lin, Ye Wu, Mengying Yang, Xu Liu, Yizhou Liu, Qiang Ke, Qin Guo

llms agents reinforcement-learning

arXiv Computation and Language

Aug 21

RepSelect: Robust LLM Unlearning via Representation Selectivity

arXiv:2606. 17168v3 Announce Type: replace Abstract: When LLM weights are open or fine-tuning is available through an API, suppressing hazardous knowledge and tendencies is not enough: removal has to be deep enough that an adversary cannot restore it.

By Filip Sondej, Yushi Yang, Adam Mahdi

llms fine-tuning

arXiv Machine Learning

Sep 10

Can an AI Assistant Really Forget? Auditable Deletion from Addressable Memory

This paper introduces a deletion interface for a pretrained language model, measuring how effectively deleted records are removed from the model’s memory. By retrofitting a support‑vector memory gate into the global attention layers of a frozen Gemma 3, the authors show that deletions can be performed without altering weights and that the resulting state is close to a reference state that never stored the record. Experiments on 4B‑parameter models demonstrate low perplexity impact and strong evidence that deleted content is hard to recover, while larger or smaller models fail to achieve the same guarantees.

By Vishwajith Ramesh

llms

arXiv AI

Aug 18

What Does Context Compression Cost an Agent? Interaction Costs Unrevealed by Task-Completion Metrics

arXiv:2608. 16370v1 Announce Type: new Abstract: Task completion is the standard metric for evaluating context compression, yet it is incomplete: compression can increase an agent's interaction cost by forcing it to reacquire dropped state while leaving completion statistically unchanged.

By Shuyu Liu

llms agents

One email a morning, machine-written

Related stories

DRSR: Learning Set-Level Deletion Risk for Efficient Long-Horizon Agents

SHRED: Retain-Set-Free Unlearning via Self-Distillation with Logit Demotion

TRACER: Per-Tool Context Retention for LLM Agents via Consequence-Attributed Reinforcement Learning

RepSelect: Robust LLM Unlearning via Representation Selectivity

Can an AI Assistant Really Forget? Auditable Deletion from Addressable Memory

What Does Context Compression Cost an Agent? Interaction Costs Unrevealed by Task-Completion Metrics