@Azaliamirh: Check out TRACE, a new self-improvement approach where the agent identifies the missing capabilities behind its own fai…
Summary
TRACE is a new self-improvement approach where an AI agent identifies the missing capabilities behind its own failures and trains itself to address them. TRACE-trained Qwen3.6-27B achieves 73.2% on SWE-bench Verified, outperforming much larger models with fewer training rollouts.
View Cached Full Text
Cached at: 07/10/26, 12:11 PM
Check out TRACE, a new self-improvement approach where the agent identifies the missing capabilities behind its own failures and trains itself to address them.
TRACE-trained Qwen3.6-27B reaches 73.2% on SWE-bench Verified, outperforming much larger models like Codex 5.2 and GLM 5, while beating GRPO and GEPA with <1/4 the training rollouts.
By contrasting successful and failed trajectories, TRACE identifies its own weaknesses (such as bug localization or retrieval of the correct doc) and creates new synthetic environments to fix them.
The result is a transferable and sample-efficient synthetic env / data generation + fine-tuning pipeline for agentic tasks.
Great work led by @TarunSures41845 and @hangoo_kang!
Hangoo Kang @ ICML ✈️ (@hangoo_kang): “TRACE: Capability-Targeted Agentic Training” got Spotlight @ ICML AIWILD 🎉
Beats direct RL, GEPA, & synthetic-agent data on SWE-Bench Verified and τ²-Bench.
TRACE-Qwen3.6-27B tops GPT-5.2-Codex, GLM 5, & Claude 4.5 Sonnet on SWE-Bench.
Co-led with @TarunSures41845. Thanks to
Similar Articles
@ADarmouni: https://arxiv.org/pdf/2607.13988 Good RL work from Microsoft research that manages to improve the Qwen3 small MoE bette…
This paper introduces TRACE, a dense credit-assignment method for reinforcement learning of long-horizon agents that significantly improves Qwen3 small MoE models on agentic benchmarks without additional critic models.
@Vtrivedy10: https://x.com/Vtrivedy10/status/2066571435871551655
A joint study by LangChain Labs and Fireworks AI demonstrates fine-tuning an open Qwen model to create a trace judge that detects 'perceived error' in production traces, achieving frontier performance at up to 100x lower cost. The model is evaluated on two internal datasets and shows generality across applications.
BenchTrace: A Benchmark for Testing Reflection Ability and Controlled Evolution in LLM Agents
BenchTrace is a benchmark for evaluating the self-evolution abilities of LLM agents, focusing on reflection and controlled evolution through a dataset of 1,821 annotated episodes and two evaluation tasks: Reflection Evaluation and Evolution Evaluation. Experiments with Qwen3-32B and GPT-4.1 show both models struggle, with a main bottleneck in diagnosis and issues in generalization and forgetting.
Qwen3.7: The Agent Frontier (15 minute read)
Alibaba's Qwen team has released Qwen3.7-Max, a proprietary agent-foundation model achieving top scores on multiple benchmarks including Terminal-Bench 2.0, SWE-Pro, and GPQA Diamond, with consistent performance across various code environments.
@IntuitMachine: Here's the local LLM solution to Satya Nadell's Reverse Information Paradox. Your AI agent is failing for the same reas…
TRACE is a method that uses contrastive diagnosis to identify an AI agent's few missing capabilities, then trains tiny LoRA adapters on synthetic micro-environments, achieving 15+ point gains on coding benchmarks with dramatically less compute.