mlsanityData to 7 Oct 2026

On Catastrophic Inheritance of Large Foundation Models

arXiv 2402.01909 ↗First posted 2 Feb 2024cs.LG cs.AI cs.CYcited by 15

Not on the map. Why?

Abstract (arXiv)

Large foundation models (LFMs) are claiming incredible performances. Yet great concerns have been raised about their mythic and uninterpreted potentials not only in machine learning, but also in various other disciplines. In this position paper, we propose to identify a neglected issue deeply rooted in LFMs: Catastrophic Inheritance, describing the weaknesses and limitations inherited from biased large-scale pre-training data to behaviors of LFMs on the downstream tasks, including samples that are corrupted, long-tailed, noisy, out-of-distributed, to name a few. Such inheritance can potentially cause catastrophes to downstream applications, such as bias, lack of generalization, deteriorated performance, security vulnerability, privacy leakage, and value misalignment. We discuss the challenges behind this issue and propose UIM, a framework to Understand the catastrophic inheritance of LFMs from both pre-training and downstream adaptation, Interpret the implications of catastrophic inheritance on downstream tasks, and how to Mitigate it. UIM aims to unite both the machine learning and social sciences communities for more responsible and promising AI development and deployment.

Key facts. Cited by 15 papers in the map (data through 7 Oct 2026). Builds on 163 of its 207 references that are in the map. First posted on arXiv 2 Feb 2024.

Builds on

163 of 207 references.

References · 163

  1. 2017Attention Is All You Needcited by 59,932
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  3. 2020Language Models are Few-Shot Learnerscited by 32,199
  4. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  5. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  6. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  7. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  8. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  9. 2023Visual Instruction Tuningcited by 7,792
  10. 2020Scaling Laws for Neural Language Modelscited by 6,411
  11. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,131
  12. 2021On the Opportunities and Risks of Foundation Modelscited by 3,961
38 more
  1. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,500
  2. 2023Mistral 7Bcited by 3,065
  3. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 2,914
  4. 2022Emergent Abilities of Large Language Modelscited by 2,483
  5. 2020The Pile: An 800GB Dataset of Diverse Text for Language Modelingcited by 2,346
  6. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,842
  7. 2022Large Language Models Encode Clinical Knowledgecited by 1,807
  8. 2016Understanding deep learning requires rethinking generalizationcited by 1,657
  9. 2023Jailbroken: How Does LLM Safety Training Fail?cited by 1,565
  10. 2021LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairscited by 1,544
  11. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
  12. 2019Do ImageNet Classifiers Generalize to ImageNet?cited by 1,507
  13. 2021Calibrate Before Use: Improving Few-Shot Performance of Language Modelscited by 1,371
  14. 2023Jailbreaking Black Box Large Language Models in Twenty Queriescited by 1,326
  15. 2020Making Pre-trained Language Models Better Few-shot Learnerscited by 1,255
  16. 2023A Survey on Evaluation of Large Language Modelscited by 1,227
  17. 2022Reproducible scaling laws for contrastive language-image learningcited by 1,163
  18. 2019A Comprehensive Survey on Transfer Learningcited by 1,154
  19. 2023PaLM 2 Technical Reportcited by 1,154
  20. 2023Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!cited by 1,106
  21. 2022When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memoriescited by 1,103
  22. 2019Machine Unlearningcited by 1,020
  23. 2021Towards a Unified View of Parameter-Efficient Transfer Learningcited by 863
  24. 2019Big Transfer (BiT): General Visual Representation Learningcited by 766
  25. 2022Quantifying Memorization Across Neural Language Modelscited by 755
  26. 2023The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Onlycited by 755
  27. 2023CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquingcited by 751
  28. 2023Extracting Training Data from Diffusion Modelscited by 723
  29. 2024Eyes Wide Shut? Exploring the Visual Shortcomings of Multimodal LLMscited by 719
  30. 2021Deduplicating Training Data Makes Language Models Bettercited by 705
  31. 2023Time-LLM: Time Series Forecasting by Reprogramming Large Language Modelscited by 680
  32. 2019Deep Double Descent: Where Bigger Models and More Data Hurtcited by 667
  33. 2023Bias and Fairness in Large Language Models: A Surveycited by 626
  34. 2022Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasetscited by 599
  35. 2023Textbooks Are All You Needcited by 561
  36. 2022Beyond neural scaling laws: beating power law scaling via data pruningcited by 519
  37. 2023Are Emergent Abilities of Large Language Models a Mirage?cited by 515
  38. 2020What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimationcited by 492

50 most cited of 163 All 163 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 10; 2025: 4; 2026: 1.

15 papers.

Sort:

By area:

  1. 2024SEAL: Safety-enhanced Aligned LLM Fine-tuning via Bilevel Data Selectioncited by 64
  2. 2024ControlVAR: Exploring Controllable Visual Autoregressive Modelingcited by 58
  3. 2024On the Diversity of Synthetic Data and its Impact on Training Large Language Modelscited by 56
  4. 2024Enhancing Zero-Shot Vision Models by Label-Free Prompt Distribution Learning and Bias Correctingcited by 26
  5. 2024As Firm As Their Foundations: Can open-sourced foundation models be used to create adversarial examples for downstream tasks?cited by 11
  6. 2024Slight Corruption in Pre-training Data Makes Better Diffusion Modelscited by 9
  7. 2025Understanding and Mitigating Bias Inheritance in LLM-based Data Augmentation on Downstream Taskscited by 9
  8. 2024Evaluating Subword Tokenization: Alien Subword Composition and OOV Generalization Challengecited by 8
  9. 2024BA-LoRA: Bias-Alleviating Low-Rank Adaptation to Mitigate Catastrophic Inheritance in Large Language Modelscited by 7
  10. 2025On Fairness of Unified Multimodal Large Language Model for Image Generationcited by 6
  11. 2025Targeted Forgetting of Image Subgroups in CLIP Modelscited by 6
  12. 2025Decoupling Template Bias in CLIP: Harnessing Empty Prompts for Enhanced Few-Shot Learningcited by 4
3 more
  1. 2024Open-Set Video-based Facial Expression Recognition with Human Expression-sensitive Promptingcited by 3
  2. 2026UrbanFM: Scaling Urban Spatio-Temporal Foundation Modelscited by 1
  3. 2024EchoAlign: Bridging Generative and Discriminative Learning under Noisy Labelscited by 0