Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

PACE: Two-Timescale Self-Evolution for Small Language Model Agents

About

Deploying language-model agents in production often requires substantial compute and human effort to tune prompts, parsers, validators, and other components of the agent pipeline. Self-evolution offers a promising alternative, but most existing frameworks assume access to frontier models that can reliably diagnose failures, propose revisions, and judge their own updates. We study whether frozen small language models (SLMs) can serve as effective self-evolving agents under resource constraints. We propose PACE (Prompt And Control Logic Evolution), a two-timescale framework that coordinates low-risk prompt refinement with higher-risk control-logic updates. PACE evolves prompts under fixed control logic until prompt-level gains saturate, then considers constrained control-logic updates that are accepted through held-out validation. Across three frozen SLM backbones ranging from 4B to 14B parameters and four controlled benchmarks, PACE achieves the best performance on all 12 backbone--benchmark combinations, improving over vanilla SLM agents by up to +9.2% relative improvement and over the stronger single-mode evolution baseline by up to +5.4% relative improvement. A tau-bench case study further shows that PACE improves multi-turn tool-use success over vanilla and prompt-only evolution. These results suggest that reliable SLM agent self-evolution is possible without updating model weights or relying on frontier-model teachers, and that the key benefit is not any single final solver pattern but autonomous, validated discovery of task-appropriate inference strategies.

Chen Ling, Pei Chen, Albert Guan, Jiaming Qu, Shayan Ali Akbar, Madhu Gopinathan, Erwin Cornejo• 2026

Related benchmarks

TaskDatasetResultRank
Mathematical ReasoningMGSM
Accuracy90.9
236
Agentic Tool-usetau2-bench Retail
Pass@181.7
30
Agentic Tool-usetau2-bench Airline
Pass@157.5
30
Multi-hop Question AnsweringHotpotQA
Score80.3
25
Instruction FollowingIFEval
IFEval Score83.7
25
Language UnderstandingMMLU
MMLU Score88.9
25
Showing 6 of 6 rows

Other info

Follow for update