Skip to content
OpenTrain AIFor AI Companies

HFEPX · Eval paper review

JudgeMix: Judge-Guided Mixed-Rollout Data Synthesis for LLM Agents

Cheng-Yu Hsieh, Ting-Yao Hu, Sanjoy Chowdhury, Oncel Tuzel +1 more

Published

Nov 12, 2026

Citations

0

Trust level

Low

Usefulness score

25/100 (Low)

Extraction confidence

45% (Low)

Derived from extracted protocol signals and abstract evidence.

Rater population

Not reported

Signals refreshed

Nov 12, 2026

Open arXiv Find implementation Shortlist (0) RSS feed

Tags

Automatic Metrics General Long Horizon Tool Use

Should you rely on this paper?

This paper is adjacent to HFEPX scope and is best used for background context, not as a primary protocol reference.

Use this as background context only. Do not make protocol decisions from this page alone.

Best use

Background context only

Use if you need

A secondary eval reference to pair with stronger protocol papers.

What to verify

Validate the evaluation procedure and quality controls in the full paper before operational use.

Main weakness

The available metadata is too thin to trust this as a primary source.

Human feedback signal
Not explicit
Not explicit in abstract metadata
Evaluation signal
Detected
Eval setup described
Usefulness for eval research
25/100
Adjacent candidate

Treat as adjacent context, not a core eval-method reference.

If you are doing eval pipeline work, start here

Human Eval Papers LLM-as-Judge Papers Pairwise Preference Papers Tool-Use Evaluation Search

Abstract

Supervised fine-tuning (SFT) on expert trajectories is a standard paradigm for training LLM agents. However, it suffers from a well-known training-inference distribution mismatch, calling for target-aware data synthesis strategies that ground the training data in the target model's native state distribution. To bridge this gap, emerging approaches incorporate the expert either to guide mixed rollouts alongside the target model, or to provide dense step-wise supervision. Nonetheless, existing methods apply expert assistance blind to the target model's inherent behavior: they either disrupt the rollout using sub-optimal random intervention heuristics, or they enforce indiscriminate uniform supervision that unnecessarily overwrites the model's already-correct actions. We introduce JudgeMix, a judge-guided data synthesis framework that dynamically interleaves the target model and expert based on step-wise semantic evaluations. JudgeMix queries expert guidance precisely when the target model struggles, ensuring high trajectory success rates while keeping the state distribution aligned with the target model. It additionally uses the judge evaluation to reweight expert supervision on different steps, preserving model's inherent valid behavior while correcting its errors. On two agentic tool-use benchmarks, AppWorld and AutomationBench, JudgeMix demonstrates consistent superior performance to existing methods, with improvements up to 10 pass@1 points.

What we could verify

These are the protocol signals we could actually recover from the available paper metadata. Use them to decide whether this paper is worth deeper reading.

Human Feedback Types

missing

None explicit

No explicit feedback protocol extracted.

"Supervised fine-tuning (SFT) on expert trajectories is a standard paradigm for training LLM agents."

Evaluation Modes

partial

Automatic Metrics

Includes extracted eval setup.

"Supervised fine-tuning (SFT) on expert trajectories is a standard paradigm for training LLM agents."

Quality Controls

missing

Not reported

No explicit QC controls found.

"Supervised fine-tuning (SFT) on expert trajectories is a standard paradigm for training LLM agents."

Benchmarks / Datasets

missing

Not extracted

No benchmark anchors detected.

"Supervised fine-tuning (SFT) on expert trajectories is a standard paradigm for training LLM agents."

Reported Metrics

partial

Pass@1

Useful for evaluation criteria comparison.

"On two agentic tool-use benchmarks, AppWorld and AutomationBench, JudgeMix demonstrates consistent superior performance to existing methods, with improvements up to 10 pass@1 points."

Benchmarks and datasets

No benchmark or dataset names were extracted from the available abstract.

Reported metrics

pass@1
Human feedback details
Uses human feedback
No
Feedback types
None
Rater population
Not reported
Unit of annotation
Trajectory (inferred)
Expertise required
General
Evaluation details
Evaluation modes
Automatic Metrics
Agentic eval
Tool Use, Long Horizon
Quality controls
Not reported
Evidence quality
Low
Use this page as
Background context only

Research brief

Metadata summary

Supervised fine-tuning (SFT) on expert trajectories is a standard paradigm for training LLM agents.

Based on abstract + metadata only. Check the source paper before making high-confidence protocol decisions.

Key takeaways

  • Supervised fine-tuning (SFT) on expert trajectories is a standard paradigm for training LLM agents.
  • However, it suffers from a well-known training-inference distribution mismatch, calling for target-aware data synthesis strategies that ground the training data in the target model's native state distribution.
  • To bridge this gap, emerging approaches incorporate the expert either to guide mixed rollouts alongside the target model, or to provide dense step-wise supervision.

Researcher actions

  • Compare this paper against nearby papers in the same arXiv category before using it for protocol decisions.
  • Check the full text for explicit evaluation design choices (raters, protocol, and metrics).
  • Use related-paper links to find stronger protocol-specific references.

Caveats

  • Generated from abstract + metadata only; no PDF parsing.
  • Signals below are heuristic and may miss details reported outside the abstract.

Recommended queries

Expert verification evaluation

Contribution summary

  • Supervised fine-tuning (SFT) on expert trajectories is a standard paradigm for training LLM agents.
  • However, it suffers from a well-known training-inference distribution mismatch, calling for target-aware data synthesis strategies that ground the training data in the target model's native state distribution.

Why it matters for eval

  • We introduce JudgeMix, a judge-guided data synthesis framework that dynamically interleaves the target model and expert based on step-wise semantic evaluations.

Researcher checklist

  • Human feedback protocol is explicit

    No explicit human feedback protocol detected.

  • Evaluation mode is explicit

    Detected: Automatic Metrics

  • Quality control reporting appears

    No calibration/adjudication/IAA control explicitly detected.

  • Benchmark or dataset anchors are present

    No benchmark/dataset anchor extracted from abstract.

  • Metric reporting is present

    Detected: pass@1