Skip to content
OpenTrain AIFor AI Companies

HFEPX · Eval paper review

Distilling the Shape of Uncertainty: What On-Policy Distillation Learns and How to Predict It

Shiping Yang, Chuanqi Tang, Hossein Hajimirsadeghi, Layla El Asri +2 more

Published

Oct 28, 2026

Citations

0

Trust level

Low

Usefulness score

40/100 (Low)

Extraction confidence

45% (Low)

Derived from extracted protocol signals and abstract evidence.

Rater population

Not reported

Signals refreshed

Oct 28, 2026

Open arXiv Find implementation Shortlist (0) RSS feed

Tags

Coding Pairwise Preference

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

Background context only.

What to verify

Read the full paper before copying any benchmark, metric, or protocol choices.

Main weakness

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

Human feedback signal
Detected
From extracted signals
Evaluation signal
Weak or implicit
Validate from full paper
Usefulness for eval research
40/100
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

On-policy distillation (OPD) is increasingly adopted in the post-training of large language models. However, a stronger teacher may fail to improve the student. Prior work attributes this to a lack of new knowledge, but what that knowledge is remains unclear. We show that OPD learns not only \textit{how to branch}, i.e., which candidate tokens to consider, but also \textit{where to branch}, i.e., where to keep reasoning paths open. The existing token-overlap metric captures only the former. We introduce the entropy shape, the sequence of next-token entropies, to characterize the latter, and propose the Entropy Shape Correlation (ESC) to measure student-teacher agreement in entropy shape. Tracking ESC throughout training and intervening on a teacher's entropy while preserving its token preferences, we empirically show that \textit{where to branch} causally drives OPD gains, independently of \textit{how to branch}. Finally, we formulate OPD outcome prediction as a teacher selection task and establish an evaluation protocol for it, under which pre-OPD ESC selects the best teacher for both students and outperforms existing predictors. Code is available at https://github.com/maybenotime/opd-esc.

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

partial

Pairwise Preference

Directly usable for protocol triage.

"On-policy distillation (OPD) is increasingly adopted in the post-training of large language models."

Evaluation Modes

missing

None explicit

Validate eval design from full paper text.

"On-policy distillation (OPD) is increasingly adopted in the post-training of large language models."

Quality Controls

missing

Not reported

No explicit QC controls found.

"On-policy distillation (OPD) is increasingly adopted in the post-training of large language models."

Benchmarks / Datasets

missing

Not extracted

No benchmark anchors detected.

"On-policy distillation (OPD) is increasingly adopted in the post-training of large language models."

Reported Metrics

missing

Not extracted

No metric anchors detected.

"On-policy distillation (OPD) is increasingly adopted in the post-training of large language models."

Benchmarks and datasets

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

Reported metrics

No metric terms were extracted from the available abstract.

Human feedback details
Uses human feedback
Yes
Feedback types
Pairwise Preference
Rater population
Not reported
Expertise required
Coding
Evaluation details
Evaluation modes
None
Agentic eval
None
Quality controls
Not reported
Evidence quality
Low
Use this page as
Background context only

Research brief

Metadata summary

On-policy distillation (OPD) is increasingly adopted in the post-training of large language models.

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

Key takeaways

  • On-policy distillation (OPD) is increasingly adopted in the post-training of large language models.
  • However, a stronger teacher may fail to improve the student.
  • Prior work attributes this to a lack of new knowledge, but what that knowledge is remains unclear.

Researcher actions

  • Compare this paper against nearby papers in the same arXiv category before using it for protocol decisions.
  • Validate inferred eval signals (Automatic metrics) against the full paper.
  • 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

Agreement / Kappa

Contribution summary

  • We show that OPD learns not only how to branch, i.e., which candidate tokens to consider, but also where to branch, i.e., where to keep reasoning paths open.
  • We introduce the entropy shape, the sequence of next-token entropies, to characterize the latter, and propose the Entropy Shape Correlation (ESC) to measure student-teacher agreement in entropy shape.
  • Tracking ESC throughout training and intervening on a teacher's entropy while preserving its token preferences, we empirically show that where to branch causally drives OPD gains, independently of how to branch.

Why it matters for eval

  • Tracking ESC throughout training and intervening on a teacher's entropy while preserving its token preferences, we empirically show that where to branch causally drives OPD gains, independently of how to branch.
  • Finally, we formulate OPD outcome prediction as a teacher selection task and establish an evaluation protocol for it, under which pre-OPD ESC selects the best teacher for both students and outperforms existing predictors.

Researcher checklist

  • Human feedback protocol is explicit

    Detected: Pairwise Preference

  • Evaluation mode is explicit

    No clear evaluation mode extracted.

  • 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

    No metric terms extracted.