# The paper argues OPD learns where reasoning paths stay open [Canonical reading](https://escapemeridian.com/readings/the-paper-argues-opd-learns-where-reasoning-paths-stay-open-e2babda5) Authors Shiping Yang, Chuanqi Tang, Hossein Hajimirsadeghi, Layla El Asri, Mohammad Amin Shabani, and Angel X. Chang argue that on-policy distillation (OPD) learns more than which candidate tokens a student considers. It also learns where the student keeps reasoning paths open, which the authors call the “shape” of uncertainty. The authors introduce entropy shape—the sequence of next-token entropies—and Entropy Shape Correlation (ESC), a measure of student-teacher agreement on that sequence. They report that intervening on a teacher’s entropy while preserving its token preferences indicates that where to branch causally drives OPD gains, independently of how to branch. They also frame predicting OPD outcomes as a teacher-selection task. In their evaluation protocol, pre-OPD ESC selects the best teacher for both students and outperforms existing predictors. If this measure helps identify a useful teacher before distillation, it could make teacher selection more informed; the abstract does not establish how well the result transfers beyond that protocol. What results would show whether pre-OPD ESC predicts teacher quality for students and settings outside the authors’ evaluation protocol? Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims. ## Original source arXiv cs.AI · Article Original source date: Oct 28, 2026 (2026-10-28). Date basis: Date basis not provided. Added to Escape Meridian: 2026-11-11T08:25:25.106Z. First observed: 2026-10-30T06:42:44.412Z. Retrieved: 2026-10-31T11:46:50.280Z. - [Distilling the Shape of Uncertainty: What On-Policy Distillation Learns and How to Predict It](https://arxiv.org/abs/2610.39020) ## Limits The evidence is the paper’s abstract; it does not provide the evaluation protocol’s details, measurements, or the scope of students and settings tested. The abstract does not establish whether pre-OPD ESC predicts teacher quality beyond the authors’ evaluation protocol. ## Open question What results would show whether pre-OPD ESC predicts teacher quality for students and settings outside the authors’ evaluation protocol? ## Source-stated The authors argue that OPD learns both which candidate tokens to consider and where to keep reasoning paths open. > 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. - [Supporting source](https://arxiv.org/abs/2610.39020) ## Source-stated The authors introduce entropy shape and ESC to characterize and measure agreement on where reasoning paths remain 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. - [Supporting source](https://arxiv.org/abs/2610.39020) ## Source-stated The authors report that where to branch causally drives OPD gains independently of how to branch, based on tracking ESC and intervening on teacher entropy while preserving token preferences. > 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}. - [Supporting source](https://arxiv.org/abs/2610.39020) ## Source-stated Under the authors’ evaluation protocol, pre-OPD ESC selects the best teacher for both students and outperforms existing predictors. > under which pre-OPD ESC selects the best teacher for both students and outperforms existing predictors. - [Supporting source](https://arxiv.org/abs/2610.39020) ## Source and contributor context Yang, Shiping, Tang, Chuanqi, Hajimirsadeghi, Hossein, Asri, Layla El, Shabani, Mohammad Amin, Chang, Angel X. · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media No suitable media was found by the supported page extraction. ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## arXiv arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed. - [Explore this context](/entities/arxiv) - [Reference](https://arxiv.org/) - [Reference](https://info.arxiv.org/help/api/index.html) ## The Sequence The Sequence publishes machine-learning analysis and interviews whose guest answers retain their own attribution. - [Explore this context](/entities/the-sequence) - [Reference](https://thesequence.substack.com/) - [Reference](https://thesequence.substack.com/p/the-sequence-chat-emad-mostaque-stability) ## Editorial provenance Version 1. Publication ID: daadfca5-7868-4fbd-8892-e560800eca6c. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-11T08:25:25.106Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: material\_development. Observation: 4858c3e836a9c21eb1834f6ebced66f1b13a4ed4b8497895227608cc71448934. Source text hash: 7d0962984b7e69e4c87e6ca815ffb254ae30c6566ba55c1ff1dc4758024ece5c. ## Revision and source-check history ### Version 1 — 2026-11-11T08:25:25.106Z Authors Shiping Yang, Chuanqi Tang, Hossein Hajimirsadeghi, Layla El Asri, Mohammad Amin Shabani, and Angel X. Chang argue that on-policy distillation (OPD) learns more than which candidate tokens a student considers. It also learns where the student keeps reasoning paths open, which the authors call the “shape” of uncertainty. The authors introduce entropy shape—the sequence of next-token entropies—and Entropy Shape Correlation (ESC), a measure of student-teacher agreement on that sequence. They report that intervening on a teacher’s entropy while preserving its token preferences indicates that where to branch causally drives OPD gains, independently of how to branch. They also frame predicting OPD outcomes as a teacher-selection task. In their evaluation protocol, pre-OPD ESC selects the best teacher for both students and outperforms existing predictors. If this measure helps identify a useful teacher before distillation, it could make teacher selection more informed; the abstract does not establish how well the result transfers beyond that protocol. What results would show whether pre-OPD ESC predicts teacher quality for students and settings outside the authors’ evaluation protocol? ## Original source arXiv cs.AI · Article Original source date: Oct 28, 2026 (2026-10-28). Date basis: Date basis not provided. Added to Escape Meridian: 2026-11-11T08:25:25.106Z. First observed: 2026-10-30T06:42:44.412Z. Retrieved: 2026-10-31T11:46:50.280Z. - [Distilling the Shape of Uncertainty: What On-Policy Distillation Learns and How to Predict It](https://arxiv.org/abs/2610.39020) ## Limits The evidence is the paper’s abstract; it does not provide the evaluation protocol’s details, measurements, or the scope of students and settings tested. The abstract does not establish whether pre-OPD ESC predicts teacher quality beyond the authors’ evaluation protocol. ## Open question What results would show whether pre-OPD ESC predicts teacher quality for students and settings outside the authors’ evaluation protocol? ## Source-stated The authors argue that OPD learns both which candidate tokens to consider and where to keep reasoning paths open. > 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. - [Supporting source](https://arxiv.org/abs/2610.39020) ## Source-stated The authors introduce entropy shape and ESC to characterize and measure agreement on where reasoning paths remain 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. - [Supporting source](https://arxiv.org/abs/2610.39020) ## Source-stated The authors report that where to branch causally drives OPD gains independently of how to branch, based on tracking ESC and intervening on teacher entropy while preserving token preferences. > 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}. - [Supporting source](https://arxiv.org/abs/2610.39020) ## Source-stated Under the authors’ evaluation protocol, pre-OPD ESC selects the best teacher for both students and outperforms existing predictors. > under which pre-OPD ESC selects the best teacher for both students and outperforms existing predictors. - [Supporting source](https://arxiv.org/abs/2610.39020) ## Source and contributor context Yang, Shiping, Tang, Chuanqi, Hajimirsadeghi, Hossein, Asri, Layla El, Shabani, Mohammad Amin, Chang, Angel X. · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media No suitable media was found by the supported page extraction. ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## arXiv arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed. - [Explore this context](/entities/arxiv) - [Reference](https://arxiv.org/) - [Reference](https://info.arxiv.org/help/api/index.html) ## The Sequence The Sequence publishes machine-learning analysis and interviews whose guest answers retain their own attribution. - [Explore this context](/entities/the-sequence) - [Reference](https://thesequence.substack.com/) - [Reference](https://thesequence.substack.com/p/the-sequence-chat-emad-mostaque-stability) ## Editorial provenance Version 1. Publication ID: daadfca5-7868-4fbd-8892-e560800eca6c. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-11T08:25:25.106Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: material\_development. Observation: 4858c3e836a9c21eb1834f6ebced66f1b13a4ed4b8497895227608cc71448934. Source text hash: 7d0962984b7e69e4c87e6ca815ffb254ae30c6566ba55c1ff1dc4758024ece5c. ### uncertain — 2026-11-11T08:25:25.106Z — anthropic/claude-opus-5.5 Publication: daadfca5-7868-4fbd-8892-e560800eca6c attribution — pass: The six listed authors match the arXiv page. The claims are framed as what the authors argue, introduce or report, not as established fact or EM opinion. > Authors:Shiping Yang, Chuanqi Tang, Hossein Hajimirsadeghi, Layla El Asri, Mohammad Amin Shabani, Angel X. Chang summary — uncertain: The reading tracks the abstract faithfully, but a non-specialist may not follow it. It never says that OPD is a post-training method in which a student language model learns from a teacher model. It also leaves 'tokens' and 'next-token entropies' unexplained (add tooltips). It omits the motivating puzzle that a stronger teacher can fail to improve the student. 'Both students' appears without saying two student models were tested. Smallest repair: in the first paragraph, add the post-training context and the stronger-teacher puzzle. Tooltips should cover OPD, token and entropy. > 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. date — pass: originalPublished 2026-10-28 matches the v1 submission date. The later feed date (Oct 31) was correctly not used. > \[Submitted on 28 Oct 2026\] support — pass: Each claim is backed by an exact abstract excerpt. The causal finding is attributed as the authors' empirical claim. The teacher-selection result is scoped to their evaluation protocol. The benefit statement is explicitly conditional, and the limitations note that the evidence is abstract-only. The question follows from that unresolved scope. > 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. media — pass: The selected media is the verified direct arXiv abstract page for this paper, which is the most direct access to the source. ## Recorded references Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter. No public references recorded here yet. [Reference data (JSON)](/readings/e2babda51d94af28b41b581a7d73079d9cc83c7e82c3ef7dda44dc18e661bfd0/references.json) ## Corrections and challenges [Send a correction or an opposing source](https://escapemeridian.com/corrections/) [JSON record](https://escapemeridian.com/readings/the-paper-argues-opd-learns-where-reasoning-paths-stay-open-e2babda5.json)