SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.04651 · 2026

SAFE: Stable Alignment Finetuning with Entropy-Aware Predictive Control for RLHF Combining Double Soft-Min Critics with Adaptive KL Thresholding

Dipan Maity

arXiv · PDF · Open in the Atlas

Code that ran

7 of 10 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (7 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 7 have no contract verdict; 0 failures of our instrument; 3 no recorded run), measured 2026-11-21; 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/2602.04651.json · badge: /api/ran/2602.04651.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
ryyzn9/SAFE community
Papers with Code archive
7 ran · 3 no recorded run
FunctionStatusWhere it lives
DoubleCritic Ran
ran on a synthesized input · constructed an object; no result was computed
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("348a8d18249af234")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

EntropyAwarePredictiveController Ran
ran on a synthesized input · constructed an object; no result was computed
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("97aca6b496126665")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

PIDController Ran
ran on a synthesized input · constructed an object; no result was computed
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("48c8a6b260a13f7f")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

PhaseDetector Ran
ran on a synthesized input · constructed an object; no result was computed
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("3c780e38c4b7bf11")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

RewardNormalizer Ran
ran on a synthesized input · constructed an object; no result was computed
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("1944d530269a8eda")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

S3KLQConfig Ran
ran on a synthesized input · constructed an object; no result was computed
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("0f00da5a257590d7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

SAFEConfig Ran
ran on a synthesized input · constructed an object; no result was computed
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("4067dfbd13ce4f31")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

BaseKLController Not yet run
unverified (harvested, no recorded run)
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("2aab167c40842355")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

BaseTrainer Not yet run
unverified (harvested, no recorded run)
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("5e6545e6f0ce1a83")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

SAFETrainer Not yet run
unverified (harvested, no recorded run)
ryyzn9/SAFE/safe/trainers/safe.py
pointer only (licence: NONE) · get_code("62b0dcfc9f2c25f0")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, ryyzn9/SAFE when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 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 10 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 2026-11-21.

Repositories linked to this paper

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

Abstract

Proximal Policy Optimization (PPO) has been positioned by recent literature as the canonical method for the RL part of Reinforcement Learning from Human Feedback (RLHF). PPO performs well empirically but has a heuristic motivation and handles the KL-divergence constraint used in LM-RLHF in an ad-hoc manner and suffers from reward oscillations, entropy collapse, value function drift, and sudden policy divergence that require frequent restarts and extensive hyperparameter tuning. In this paper, we develop a new pure on policy actor-critic RL method for the LM-RLHF setting. We present SAFE (Stable Alignment Finetuning with Entropy-aware control),a novel RLHF algorithm that combines a Double Soft-Min Critic for pessimistic value estimation with a new multi-layer stabilization framework combining entropygated KL regulation, and Proportional-Integral-Derivative (PID)-controlled adaptive thresholds. Unlike standard PPO's symmetric KL penalties, SAFE distinguishes high-entropy exploration from low-entropy mode collapse and adjusts penalties dynamically based on reward velocity. Experiments on a 3B parameter model show SAFE achieves +5.15% trainingaverage reward than PPO (0.725 vs 0.689), negligible reward crashes, and superior KL control than ppo . Our method adds minimal computational overhead and provides an interpretable, crash-resistant RLHF framework that maintains aggressive learning speed while ensuring stable long-horizon optimization suitable for production deployment. Code is available at https://github.com/ryyzn9/SAFE

For agents

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

get_harvested_code_for_paper("2602.04651")
get_code_for_paper("2602.04651")
have("2602.04651")

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

curl https://syntology.ai/api/ran/2602.04651.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.04651.svg)](https://syntology.ai/paper/2602.04651)

Connect an agent — have() is free.