SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2603.10359 · 2026

HEAL: Hindsight Entropy-Assisted Learning for Reasoning Distillation

Kai Wang, Shuming Shi, Yi Shen, Zhaoxiang Liu, Shiguo Lian, Ping Chen, Ning Wang, Jieyun Huang, Wenjing Zhang, Jiangze Yan

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 2 failures of our instrument; 0 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2603.10359.json · badge: /api/ran/2603.10359.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
modelscope/ms-swift community
Papers with Code archive
2 ran · 0 no recorded run
FunctionStatusWhere it lives
_sdar_agg_loss Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
modelscope/ms-swift/swift/rl_core/advantage.py
code served (permissive licence) · get_code("2b9b9cf63839d7c6")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create loss_mat: shape (batch_size=2, seq_len=4)
loss_mat = torch.randn(2, 4)

# Create loss_mask: shape (batch_size=2, seq_len=4), binary mask
loss_mask = torch.ones(2, 4)
loss_mask[0, 2:] = 0  # Mask out some tokens in first sequence

# loss_agg_mode: one of the supported modes
loss_agg_mode = 'token-mean'

ARGS = {
    'loss_mat': loss_mat,
    'loss_mask': loss_mask,
    'loss_agg_mode': loss_agg_mode,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
compute_sdar_loss Ran
ran; our contract draft was wrong, not the code
modelscope/ms-swift/swift/rl_core/advantage.py
code served (permissive licence) · get_code("8001b3c66f9ad90c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Batch size: 2, sequence length: 3
batch_size = 2
seq_length = 3

# Create student log probabilities [B, T]
student_log_probs = torch.randn(batch_size, seq_length, requires_grad=True)

# Create teacher log probabilities [B, T]
teacher_log_probs = torch.randn(batch_size, seq_length)

# Create response mask [B, T] with binary values
response_mask = torch.ones(batch_size, seq_length)

# Gate beta parameter
gate_beta = 5.0

# Loss aggregation mode
loss_agg_mode = 'token-mean'

ARGS = {
    'student_log_probs': student_log_probs,
    'teacher_log_probs': teacher_log_probs,
    'response_mask': response_mask,
    'gate_beta': gate_beta,
    'loss_agg_mode': loss_agg_mode,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 2 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Distilling reasoning capabilities from Large Reasoning Models (LRMs) into smaller models is typically constrained by the limitation of rejection sampling. Standard methods treat the teacher as a static filter, discarding complex "corner-case" problems where the teacher fails to explore valid solutions independently, thereby creating an artificial "Teacher Ceiling" for the student. In this work, we propose Hindsight Entropy-Assisted Learning (HEAL), an RL-free framework designed to bridge this reasoning gap. Drawing on the educational theory of the Zone of Proximal Development(ZPD), HEAL synergizes three core modules: (1) Guided Entropy-Assisted Repair (GEAR), an active intervention mechanism that detects critical reasoning breakpoints via entropy dynamics and injects targeted hindsight hints to repair broken trajectories; (2) Perplexity-Uncertainty Ratio Estimator (PURE), a rigorous filtering protocol that decouples genuine cognitive breakthroughs from spurious shortcuts; and (3) Progressive Answer-guided Curriculum Evolution (PACE), a three-stage distillation strategy that organizes training from foundational alignment to frontier breakthrough. Extensive experiments on multiple benchmarks demonstrate that HEAL significantly outperforms traditional SFT distillation and other baselines.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2603.10359")
get_code_for_paper("2603.10359")
have("2603.10359")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2603.10359.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2603.10359.svg)](https://syntology.ai/paper/2603.10359)

Connect an agent — have() is free.