mlsanityData to 14 Oct 2026

Bergson: An Open Source Library for Data Attribution

arXiv 2606.11660 ↗First posted 10 Jun 2026cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

Data attribution is a promising field in interpretability that aims to explain model behavior through the influence of its training data, with applications including debugging undesirable model behavior and training dataset curation. However, significant engineering effort is required to perform it at scale, and many cutting edge techniques lack open-source tooling and support. Bergson is an open source library that aims to enable faster progress in the field by providing a host of techniques that scale to very large language models and pre-training datasets. The library natively supports on-disk gradient stores and multi-node distributed training, and provides quality of life tools for researchers. Finally, we introduce the first open-source implementations of three leading data attribution methods: MAGIC, SOURCE, and TrackStar. The library is available at https://github.com/EleutherAI/bergson .

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 37 of its 59 references that are in the map. First posted on arXiv 10 Jun 2026.

Builds on

37 of 59 references.

References · 37

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  4. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  5. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,626
  6. 2022Emergent Abilities of Large Language Modelscited by 2,512
  7. 2023Pythia: A Suite for Analyzing Large Language Models Across Training and Scalingcited by 1,786
  8. 2017Understanding Black-box Predictions via Influence Functionscited by 1,741
  9. 2025Understanding R1-Zero-Like Training: A Critical Perspectivecited by 1,345
  10. 2020Estimating Training Data Influence by Tracing Gradient Descentcited by 580
  11. 2024LESS: Selecting Influential Data for Targeted Instruction Tuningcited by 532
  12. 2024The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearningcited by 509
25 more
  1. 2024WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMscited by 496
  2. 2023Studying Large Language Model Generalization with Influence Functionscited by 326
  3. 2023TRAK: Attributing Model Behavior at Scalecited by 283
  4. 2018Representer Point Selection for Explaining Deep Neural Networkscited by 187
  5. 2020Embedding-based Retrieval in Facebook Searchcited by 156
  6. 2020Explaining Black Box Predictions and Unveiling Data Artifacts through Influence Functionscited by 144
  7. 2021Scaling Up Influence Functionscited by 132
  8. 2024MATES: Model-Aware Data Selection for Efficient Pretraining with Data Influence Modelscited by 100
  9. 2021Input Similarity from the Neural Network Perspectivecited by 79
  10. 2021Manipulating SGD with Data Ordering Attackscited by 74
  11. 2024A New Perspective on Shampoo's Preconditionercited by 59
  12. 2024Scalable Influence and Fact Tracing for Large Language Model Pretrainingcited by 42
  13. 2025Deep Ignorance: Filtering Pretraining Data Builds Tamper-Resistant Safeguards into Open-Weight LLMscited by 42
  14. 2024Procedural Knowledge in Pretraining Drives Reasoning in Large Language Modelscited by 41
  15. 2024Training Data Attribution via Approximate Unrolled Differentiationcited by 34
  16. 2022ModelDiff: A Framework for Comparing Learning Algorithmscited by 30
  17. 2025Optimizing ML Training with Metagradient Descentcited by 23
  18. 2025MAGIC: Near-Optimal Data Attribution for Deep Learningcited by 22
  19. 2022An Empirical Study of Memorization in NLPcited by 21
  20. 2022Towards Tracing Factual Knowledge in Language Models Back to the Training Datacited by 21
  21. 2024Causal Estimation of Memorisation Profilescited by 20
  22. 2024textttdattri: A Library for Efficient Data Attributioncited by 18
  23. 2026Shaping capabilities with token-level data filteringcited by 13
  24. 2024SimpleFSDP: Simpler Fully Sharded Data Parallel with torch.compilecited by 7
  25. 2025Not Every Token Needs Forgetting: Selective Unlearning to Limit Change in Utility in Large Language Model Unlearningcited by 7

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026TAME: Token Attribution and Masking for Emergent misalignmentcited by 1
  2. 2026Capability Provenance in Language Models: A Case Study in Social Reasoningcited by 0
  3. 2026dattri-LLM: A Unified and Efficient Library for Training Data Attribution at LLM Scalecited by 0