specificityplus

👩‍💻 Code for the ACL paper "Detecting Edit Failures in LLMs: An Improved Specificity Benchmark"

https://github.com/apartresearch/specificityplus

Science Score: 36.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
    Found codemeta.json file
  • .zenodo.json file
    Found .zenodo.json file
  • DOI references
  • Academic publication links
  • Committers with academic emails
    2 of 11 committers (18.2%) from academic institutions
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (8.8%) to scientific vocabulary

Keywords

benchmarking llm
Last synced: 6 months ago · JSON representation

Repository

👩‍💻 Code for the ACL paper "Detecting Edit Failures in LLMs: An Improved Specificity Benchmark"

Basic Info
Statistics
  • Stars: 20
  • Watchers: 2
  • Forks: 4
  • Open Issues: 2
  • Releases: 0
Topics
benchmarking llm
Created about 3 years ago · Last pushed about 2 years ago
Metadata Files
Readme License Citation

README.md

Detecting Edit Failures in LLMs: An Improved Specificity Benchmark (website)

This repository contains the code for the paper Detecting Edit Failures in LLMs: An Improved Specificity Benchmark (ACL Findings 2023).

It extends previous work on model editing by Meng et al. [1] by introducing a new benchmark, called CounterFact+, for measuring the specificity of model edits.

Attribution

The repository is a fork of MEMIT, which implements the model editing algorithms MEMIT (Mass Editing Memory in a Transformer) and ROME (Rank-One Model Editing). Our fork extends this code by additional evaluation scripts implementing the CounterFact+ benchmark. For installation instructions see the original repository.

Installation

We recommend conda for managing Python, CUDA, and PyTorch; pip is for everything else. To get started, simply install conda and run:

bash CONDA_HOME=$CONDA_HOME ./scripts/setup_conda.sh

$CONDA_HOME should be the path to your conda installation, e.g., ~/miniconda3.

Running Experiments

See INSTRUCTIONS.md for instructions on how to run the experiments and evaluations.

How to Cite

If you find our paper useful, please consider citing as:

```bibtex

@inproceedings{jason2023detecting, title = {Detecting Edit Failures In Large Language Models: An Improved Specificity Benchmark}, author = {Hoelscher-Obermaier, Jason and Persson, Julia and Kran, Esben and Konstas, Ionnis and Barez, Fazl}, booktitle = {Findings of ACL}, year = {2023}, organization = {Association for Computational Linguistics} }

Owner

  • Name: apartresearch
  • Login: apartresearch
  • Kind: organization
  • Email: operations@apartresearch.com

GitHub Events

Total
Last Year

Committers

Last synced: about 2 years ago

All Time
  • Total Commits: 454
  • Total Committers: 11
  • Avg Commits per committer: 41.273
  • Development Distribution Score (DDS): 0.467
Past Year
  • Commits: 197
  • Committers: 7
  • Avg Commits per committer: 28.143
  • Development Distribution Score (DDS): 0.635
Top Committers
Name Email Commits
JuliaHPersson j****n@g****m 242
Jason Hoelscher-Obermaier j****r@g****m 104
- - 55
Esben Kran e****n@k****i 33
Kevin Meng k****1@g****m 6
Fazl Barez s****9@l****k 3
Jason Hoelscher-Obermaier j****o 3
fbarez 3****z 3
David Bau d****u@g****m 2
Julia Persson 5****n 2
Fazl Barez s****9@u****k 1
Committer Domains (Top 20 + Academic)

Issues and Pull Requests

Last synced: about 2 years ago

All Time
  • Total issues: 24
  • Total pull requests: 39
  • Average time to close issues: 20 days
  • Average time to close pull requests: 2 days
  • Total issue authors: 3
  • Total pull request authors: 4
  • Average comments per issue: 0.92
  • Average comments per pull request: 0.33
  • Merged pull requests: 36
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 11
  • Pull requests: 27
  • Average time to close issues: about 1 month
  • Average time to close pull requests: about 5 hours
  • Issue authors: 3
  • Pull request authors: 4
  • Average comments per issue: 1.0
  • Average comments per pull request: 0.37
  • Merged pull requests: 26
  • Bot issues: 0
  • Bot pull requests: 0
Top Authors
Issue Authors
  • jas-ho (20)
  • fbarez (3)
  • esbenkc (1)
Pull Request Authors
  • jas-ho (25)
  • JuliaHPersson (7)
  • fbarez (5)
  • esbenkc (3)
Top Labels
Issue Labels
enhancement (1)
Pull Request Labels
enhancement (1)

Dependencies

baselines/mend/requirements.txt pypi
  • allennlp *
  • click ==7.1.2
  • datasets *
  • hydra-core *
  • jsonlines *
  • numpy *
  • spacy *
  • torch *
  • wandb *