Skip to content

YangRui2015/Generalizable-Reward-Model

View on GitHub ↗Related repositories →

Code for NeurIPS 2024 paper "Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs" (under preparation)

active 2024-10-162026-01-05 (UTC)

Complete coverage26,619 / 26,619 hourly files (100%) · 2 absent upstream2023-08-152026-08-28 (UTC)
Events
131
Pushes
47
Pull requests
1
Issues
11
Stars
39
Forks
4

Activity over time

Daily event counts in the loaded window

Line chart, 447 days from 2024-10-16 to 2026-01-05. Pushes: 47 total, peak 13 in a day. Pull requests: 1 total, peak 1 in a day. Issues: 11 total, peak 2 in a day. Comments: 27 total, peak 6 in a day. Stars: 39 total, peak 4 in a day.

  • Pushes
  • Pull requests
  • Issues
  • Comments
  • Stars

Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.

Top contributors

Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity

ContributorContributionsPushesPRsComments
YangRui20154227014
ruomengd212001
zzwjames7004
Joe-Hall-Lee5002
1sir33003
Huangzisu3001
BIRlz2002
Weiww-Xu1000
glgjss9601000
brchristian1010

Recent activity

Latest issues, pull requests and releases

  • Pull request#11brchristian2025-09-03 21:44
  • Issue#9Weiww-Xu2025-06-30 10:04
    Question about indexing logic for chosen/rejected pairs
  • Issue comment#81sir32025-06-06 23:35
    Addressing the Performance Gap: LoRA-Fine-Tuned GRM Evaluation on reward-bench
  • Issue comment#81sir32025-06-06 23:33
    Addressing the Performance Gap: LoRA-Fine-Tuned GRM Evaluation on reward-bench
  • Issue comment#81sir32025-06-05 17:26
    Addressing the Performance Gap: LoRA-Fine-Tuned GRM Evaluation on reward-bench
  • Issue comment#8YangRui20152025-06-05 17:17
    Addressing the Performance Gap: LoRA-Fine-Tuned GRM Evaluation on reward-bench
  • Issue comment#7YangRui20152025-05-08 22:35
    Training details for released GRM series
  • Issue#4zzwjames2025-04-07 04:59
    do you have code to reproduce the results of reward overoptimization for PPO
  • Issue comment#4zzwjames2025-04-07 04:59
    do you have code to reproduce the results of reward overoptimization for PPO
  • Issue comment#4YangRui20152025-04-06 04:03
    do you have code to reproduce the results of reward overoptimization for PPO
  • Issue comment#4zzwjames2025-04-05 23:39
    do you have code to reproduce the results of reward overoptimization for PPO
  • Issue comment#6YangRui20152025-03-27 17:01
    Duplicate bos token
  • Issue comment#6Huangzisu2025-03-27 16:55
    Duplicate bos token
  • Issue#6Huangzisu2025-03-27 16:55
    Duplicate bos token
  • Issue#6Huangzisu2025-03-27 09:24
    Duplicate bos token
  • Issue comment#5Joe-Hall-Lee2025-03-14 10:12
    RuntimeError: size mismatch
  • Issue#5Joe-Hall-Lee2025-03-14 10:12
    RuntimeError: size mismatch
  • Issue comment#5YangRui20152025-03-13 15:16
    RuntimeError: size mismatch
  • Issue comment#4YangRui20152025-03-01 06:59
    do you have code to reproduce the results of reward overoptimization for PPO
  • Issue#4zzwjames2025-02-28 23:39
    do you have code to reproduce the results of reward overoptimization for PPO
  • Issue#2YangRui20152025-02-24 02:28
    evaluation of full finetuned model
  • Issue comment#2BIRlz2025-02-24 02:27
    evaluation of full finetuned model
  • Issue comment#3YangRui20152025-02-20 15:33
    peft_name in scripts/eval_bt_rm.sh
  • Issue#3glgjss9602025-02-20 15:02
    peft_name in scripts/eval_bt_rm.sh
  • Issue comment#2ruomengd2025-02-20 07:57
    evaluation of full finetuned model

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 39 stars here means stars gained during the window, not the repo's star count.