SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.09345 · NeurIPS · 2024

InfoRM: Mitigating Reward Hacking in RLHF via Information-Theoretic Reward Modeling

Dacheng Tao, Liang Ding, Lefei Zhang, Sen Zhang, Yuchun Miao, Rong Bao, R Encoder

arXiv · PDF · Open in the Atlas

Code that ran

5 of 9 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; 5 failures of our instrument; 4 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; 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/2402.09345.json · badge: /api/ran/2402.09345.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
miaoyuchun/InfoRM canonical (classifier label)
classifier basis: census
3 ran · 4 no recorded run
miaoyuchun/inform canonical (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
FunctionStatusWhere it lives
cal_mahalanobis Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
miaoyuchun/inform/openrlhf/models/model_inform.py
pointer only (licence: NONE) · get_code("43e3e53f8b3f61f9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/inform 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.

compute_approx_kl Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
miaoyuchun/InfoRM/openrlhf/models/utils.py
pointer only (licence: NONE) · get_code("95084b4b49902e3d")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/InfoRM 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.

kl_divergence Ran
ran; our contract draft was wrong, not the code
miaoyuchun/inform/openrlhf/models/model_inform.py
pointer only (licence: NONE) · get_code("04d7f678f5b3f532")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/inform 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.

log_probs_from_logits Ran
ran; our contract draft was wrong, not the code
miaoyuchun/InfoRM/openrlhf/models/utils.py
pointer only (licence: NONE) · get_code("0612b61574b81472")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/InfoRM 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.

strip_sequence Ran
ran; our contract draft was wrong, not the code
miaoyuchun/InfoRM/openrlhf/cli/serve_rm.py
pointer only (licence: NONE) · get_code("ca3514c46a1261d7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/InfoRM 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.

compute_reward Not yet run
unverified (harvested, no recorded run)
miaoyuchun/InfoRM/openrlhf/models/utils.py
pointer only (licence: NONE) · get_code("0d27f3b8c425d79f")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/InfoRM 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.

convert_ring_attn_params Not yet run
unverified (harvested, no recorded run)
miaoyuchun/InfoRM/openrlhf/models/ring_attn_utils.py
pointer only (licence: NONE) · get_code("d66f259aaf6cd9c7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/InfoRM 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.

pad_sequences Not yet run
unverified (harvested, no recorded run)
miaoyuchun/InfoRM/openrlhf/models/ring_attn_utils.py
pointer only (licence: NONE) · get_code("b26627fa753393a8")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/InfoRM 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.

reset_ring_attn_position_ids Not yet run
unverified (harvested, no recorded run)
miaoyuchun/InfoRM/openrlhf/models/ring_attn_utils.py
pointer only (licence: NONE) · get_code("e93117d99aa6b210")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, miaoyuchun/InfoRM 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 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 7 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 7 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.

FunctionAlso present inEarliest-datedOthers named
reset_ring_attn_position_ids 15 other papers
1 dated earlier, 14 later
2306.00398 (2023-06)
Preference-grounded Token-level Guidance for Language Model Fine-tuning
2404.18922 (2024-04)
DPO Meets PPO: Reinforced Token Optimization for RLHF
2412.01981 (2024-12)
Free Process Rewards without Process Labels
and 12 more not named here
convert_ring_attn_params 9 other papers
1 dated earlier, 8 later
2306.00398 (2023-06)
Preference-grounded Token-level Guidance for Language Model Fine-tuning
2404.18922 (2024-04)
DPO Meets PPO: Reinforced Token Optimization for RLHF
2412.01981 (2024-12)
Free Process Rewards without Process Labels
and 6 more not named here
compute_reward 6 other papers
6 later
2504.07912 (2025-04)
Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining
2504.08837 (2025-04)
VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learning
2505.15456 (2025-05)
Teaching Language Models to Evolve with Users: Dynamic Profile Modeling for Personalized Alignment
and 3 more not named here
strip_sequence 5 other papers
5 later
2502.03373 (2025-02)
Demystifying Long Chain-of-Thought Reasoning in LLMs
2502.18968 (2025-02)
Know You First and Be You Better: Modeling Human-Like User Simulators via Implicit Profiles
2504.07912 (2025-04)
Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining
and 2 more not named here
compute_approx_kl 4 other papers
4 later
2504.07912 (2025-04)
Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining
2505.15456 (2025-05)
Teaching Language Models to Evolve with Users: Dynamic Profile Modeling for Personalized Alignment
2506.08011 (2025-06)
Play to Generalize: Learning to Reason Through Game Play
and 1 more not named here
pad_sequences 4 other papers
4 later
2504.07912 (2025-04)
Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining
2505.15456 (2025-05)
Teaching Language Models to Evolve with Users: Dynamic Profile Modeling for Personalized Alignment
2506.08011 (2025-06)
Play to Generalize: Learning to Reason Through Game Play
and 1 more not named here
log_probs_from_logits 3 other papers
3 later
2504.08837 (2025-04)
VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learning
2506.08011 (2025-06)
Play to Generalize: Learning to Reason Through Game Play
2601.05073 (2026-01)
Milestones over Outcome: Unlocking Geometric Reasoning with Sub-Goal Verifiable Reward

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

Despite the success of reinforcement learning from human feedback (RLHF) in aligning language models with human values, reward hacking, also termed reward overoptimization, remains a critical challenge. This issue primarily arises from reward misgeneralization, where reward models (RMs) compute reward using spurious features that are irrelevant to human preferences. In this work, we tackle this problem from an information-theoretic perspective and propose a framework for reward modeling, namely InfoRM, by introducing a variational information bottleneck objective to filter out irrelevant information. Notably, we further identify a correlation between overoptimization and outliers in the IB latent space of InfoRM, establishing it as a promising tool for detecting reward overoptimization. Inspired by this finding, we propose the Cluster Separation Index (CSI), which quantifies deviations in the IB latent space, as an indicator of reward overoptimization to facilitate the development of online mitigation strategies. Extensive experiments on a wide range of settings and RM scales (70M, 440M, 1.4B, and 7B) demonstrate the effectiveness of InfoRM. Further analyses reveal that InfoRM's overoptimization detection mechanism is not only effective but also robust across a broad range of datasets, signifying a notable advancement in the field of RLHF. Code is available at: https://github.com/miaoyuchun/InfoRM.

For agents

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

get_harvested_code_for_paper("2402.09345")
get_code_for_paper("2402.09345")
have("2402.09345")

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

curl https://syntology.ai/api/ran/2402.09345.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.09345.svg)](https://syntology.ai/paper/2402.09345)

Connect an agent — have() is free.