Awesome
Awesome Lifelong Learning Methods for Large Language Models (Updated Regularly, Add Papers from ACL 2024)
Contents
Introduction
This repository collects awesome survey, resource, and paper for Lifelong Learning with Large Language Models. Please refer to this survey for detailed introduction. Please refer to this repository for implementation.
Survey
- Towards Lifelong Learning of Large Language Models: A Survey (arXiv 2024.06)
- Recent Advances of Foundation Language Models-based Continual Learning: A Survey (arXiv 2024.05)
- Continual Learning of Large Language Models: A Comprehensive Survey (arXiv 2024.04)
- Towards Incremental Learning in Large Language Models: A Critical Review (arXiv 2024.04)
- Continual Learning for Large Language Models: A Survey (arXiv 2024.02)
- Continual Learning with Pre-Trained Models: A Survey (arXiv 2024.01)
- A Comprehensive Survey of Continual Learning: Theory, Method and Application (TPAMI 2024)
- How Do Large Language Models Capture the Ever-changing World Knowledge? A Review of Recent Advances (EMNLP 2023)
- Revisiting Class-Incremental Learning with Pre-Trained Models: Generalizability and Adaptivity are All You Need (arXiv 2023.03)
- Deep Class-Incremental Learning: A Survey (arXiv 2023.03)
- A Comprehensive Survey of Continual Learning: Theory, Method and Application (arXiv 2023.02)
- Continual Learning of Natural Language Processing Tasks: A Survey (arXiv 2022.11)
- Continual Lifelong Learning in Natural Language Processing: A Survey (COLING 2020)
Codebase
- https://github.com/zzz47zzz/codebase-for-incremental-learning-with-llm
- https://github.com/UIC-Liu-Lab/ContinualLM
- https://github.com/aimagelab/mammoth
- https://github.com/zhoudw-zdw/RevisitingCIL
- https://github.com/G-U-N/PyCIL
- https://github.com/ZixuanKe/PyContinual
Paper
Content
Keywords Convention
Papers
2024 (2024.08-2024.10)
Coming soon!
2024 (2024.05-2024.07)
-
Unlocking Continual Learning Abilities in Language Models, Preprint 2024.
-
Large Language Model Can Continue Evolving From Mistakes, Preprint 2024.
-
D-CPT Law: Domain-specific Continual Pre-Training Scaling Law for Large Language Models, Preprint 2024.
-
COPAL: Continual Pruning in Large Language Generative Models, ICML 2024.
-
Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attention, Preprint 2024.
-
AdapterSwap: Continuous Training of LLMs with Data Removal and Access-Control Guarantees, Preprint 2024.
-
Efficient Continual Pre-training for Building Domain Specific Large Language Models, ACL (Findings) 2024.
-
Mitigating Catastrophic Forgetting in Language Transfer via Model Merging, Preprint 2024.
-
Reuse, Don’t Retrain: A Recipe for Continued Pretraining of Language Models, Preprint 2024.
-
Breaking Language Barriers: Cross-Lingual Continual Pre-Training at Scale, Preprint 2024.
-
MoE-CT: A Novel Approach For Large Language Models Training With Resistance To Catastrophic Forgetting, Preprint 2024.
-
Leitner-Guided Memory Replay for Cross-lingual Continual Learning, NAACL 2024.
-
BAMBINO-LM: (Bilingual-)Human-Inspired Continual Pretraining of BabyLM, Preprint 2024.
-
Improving Language Models Trained with Translated Data via Continual Pre-Training and Dictionary Learning Analysis, Preprint 2024.
-
Boosting Large Language Models with Continual Learning for Aspect-based Sentiment Analysis, Preprint 2024.
-
ChronosLex: Time-aware Incremental Training for Temporal Generalization of Legal Classification Tasks, ACL 2024.
-
XMC-Agent : Dynamic Navigation over Scalable Hierarchical Index for Incremental Extreme Multi-label Classification, ACL (Findings) 2024.
-
Learn or Recall? Revisiting Incremental Learning with Pre-trained Language Models, ACL 2024.
-
Flexible Weight Tuning and Weight Fusion Strategies for Continual Named Entity Recognition, ACL (Findings) 2024.
-
Distilling Causal Effect of Data in Continual Few-shot Relation Learning, LREC-COLING 2024.
-
Improving Continual Few-shot Relation Extraction through Relational Knowledge Distillation and Prototype Augmentation, LREC-COLING 2024.
-
Enhancing Contrastive Learning with Noise-Guided Attack: Towards Continual Relation Extraction in the Wild, ACL 2024.
-
An Ensemble-of-Experts Framework for Rehearsal-free Continual Relation Extraction, ACL (Findings) 2024.
-
Continual Few-shot Relation Extraction via Adaptive Gradient Correction and Knowledge Decomposition, ACL (Findings) 2024.
-
Continual Learning with Semi-supervised Contrastive Distillation for Incremental Neural Machine Translation, ACL 2024.
-
Continual Dialogue State Tracking via Reason-of-Select Distillation, ACL (Findings) 2024,
-
Interpretable Catastrophic Forgetting of Large Language Model Fine-tuning via Instruction Vector, Preprint 2024.
-
Learn it or Leave it: Module Composition and Pruning for Continual Learning, Preprint 2024.
-
Dirichlet Continual Learning: Tackling Catastrophic Forgetting in NLP, UAI 2024.
-
PromptDSI: Prompt-based Rehearsal-free Instance-wise Incremental Learning for Document Retrieval, Preprint 2024.
-
Revisiting Catastrophic Forgetting in Large Language Model Tuning, Preprint 2024.
-
HFT: Half Fine-Tuning for Large Language Models, Preprint 2024.
-
Overcoming Catastrophic Forgetting by Exemplar Selection in Task-oriented Dialogue System, ACL 2024.
-
SAPT: A Shared Attention Framework for Parameter-Efficient Continual Learning of Large Language Models, ACL 2024.
-
Self-Evolving GPT: A Lifelong Autonomous Experiential Learner, ACL 2024.
-
LoRAMoE: Alleviating World Knowledge Forgetting in Large Language Models via MoE-Style Plugin, ACL 2024.
-
Mitigate Negative Transfer with Similarity Heuristic Lifelong Prompt Tuning, ACL (Findings) 2024.
-
Q-Tuning: Queue-based Prompt Tuning for Lifelong Few-shot Language Learning, NAACL(Findings) 2024.
-
Sequential Editing for Lifelong Training of Speech Recognition Models, Preprint 2024.
-
WISE: Rethinking the Knowledge Memory for Lifelong Model Editing of Large Language Models, Preprint 2024.
-
Lifelong Knowledge Editing for LLMs with Retrieval-Augmented Continuous Prompt Learning, Preprint 2024.
-
Detoxifying Large Language Models via Knowledge Editing, ACL 2024.
-
Model Editing at Scale leads to Gradual and Catastrophic Forgetting, ACL (Findings) 2024.
-
Can We Continually Edit Language Models? On the Knowledge Attenuation in Sequential Model Editing, ACL (Findings) 2024.
-
Incremental Sequential Labeling: A Tale of Two Shifts, ACL 2024 (Findings).
-
HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models, Preprint 2024.
-
Towards Practical Tool Usage for Continually Learning LLMs, Preprint 2024.
-
LoRA Learns Less and Forgets Less, Preprint 2024.
2024 (2024.01-2024.04)
-
CorpusBrain++: A Continual Generative Pre-Training Framework for Knowledge-Intensive Language Tasks, Preprint 2024.
-
Examining Forgetting in Continual Pre-training of Aligned Large Language Models, Preprint 2024.
-
Rho-1: Not All Tokens Are What You Need, Preprint 2024.
-
SOLAR 10.7B: Scaling Large Language Models with Simple yet Effective Depth Up-Scaling, NAACL 2024.
-
LLaMA Pro: Progressive LLaMA with Block Expansion, Preprint 2024.
-
Embracing Language Inclusivity and Diversity in CLIP through Continual Language Learning, AAAI 2024.
-
Simple and Scalable Strategies to Continually Pre-train Large Language Models, Preprint 2024.
-
Set the Clock: Temporal Alignment of Pretrained Language Models, ACL (Findings) 2024.
-
HOP to the Next Tasks and Domains for Continual Learning in NLP, AAAI 2024.
-
Bayesian Parameter-Efficient Fine-Tuning for Overcoming Catastrophic Forgetting, Preprint 2024.
-
Rehearsal-Free Modular and Compositional Continual Learning for Language Models, NAACL 2024.
-
Few-shot Incremental Event Detection, TALLIP 2024.
-
Self-generated Replay Memories for Continual Neural Machine Translation, Preprint 2024.
-
F-MALLOC: Feed-forward Memory Allocation for Continual Learning in Neural Machine Translation, NAACL 2024.
-
SAPT: A Shared Attention Framework for Parameter-Efficient Continual Learning of Large Language Models, Preprint 2024.
-
InsCL: A Data-efficient Continual Learning Paradigm for Fine-tuning Large Language Models with Instructions, NAACL 2024.
-
Analyzing and Reducing Catastrophic Forgetting in Parameter Efficient Tuning, Preprint 2024.
-
Mitigating Catastrophic Forgetting in Large Language Models with Self-Synthesized Rehearsal, Preprint 2024.
-
Scalable Language Model with Generalized Continual Learning, ICLR 2024.
-
Q-Tuning: Queue-based Prompt Tuning for Lifelong Few-shot Language, NAACL (Findings) 2024.
-
MoRAL: MoE Augmented LoRA for LLMs' Lifelong Learning, Preprint 2024.
-
WilKE: Wise-Layer Knowledge Editor for Lifelong Knowledge Editing, Preprint 2024.
-
CPPO: Continual Learning for Reinforcement Learning with Human Feedback, ICLR 2024.
-
COPR: Continual Human Preference Learning via Optimal Policy Regularization, Preprint 2024.
-
Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection, ICLR 2024.
-
LLMs in the Imaginarium: Tool Learning through Simulated Trial and Error, Preprint 2024.
-
Towards Practical Tool Usage for Continually Learning LLMs, Preprint 2024.
-
Confucius: Iterative Tool Learning from Introspection Feedback by Easy-to-Difficult Curriculum, AAAI 2024.
-
Toolllm: Facilitating large language models to master 16000+ real-world apis, ICLR 2024.
2023
-
Continual Pre-Training of Large Language Models: How to (re)warm your model?, ICML (Workshop) 2023.
-
EcomGPT-CT: Continual Pre-training of E-commerce Large Language Models with Semi-structured Data, Preprint 2023.
-
QUERT: Continual Pre-training of Language Model for Query Understanding in Travel Domain Search, KDD 2023.
-
Recyclable Tuning for Continual Pre-training, ACL (Findings) 2023.
-
Large Language Models Encode Clinical Knowledge, Nature 2023.
-
Continual Learning Under Language Shift, Preprint 2023.
-
Exploring Continual Learning for Code Generation Models, ACL 2023.
-
Lifelong Language Pretraining with Distribution-Specialized Experts, ICML 2023.
-
Mitigating Catastrophic Forgetting in Task-Incremental Continual Learning with Adaptive Classification Criterion, Preprint 2023.
-
Rehearsal-free Continual Language Learning via Efficient Parameter Isolation, ACL 2023.
-
Class-Incremental Learning based on Label Generation, ACL 2023.
-
Overcoming Catastrophic Forgetting in Massively Multilingual Continual Learning, ACL (Findings) 2023.
-
InfoCL: Alleviating Catastrophic Forgetting in Continual Text Classification from An Information Theoretic Perspective, EMNLP 2023.
-
ConPET: Continual Parameter-Efficient Tuning for Large Language Models, Preprint 2023.
-
A Neural Span-Based Continual Named Entity Recognition Model, AAAI 2023.
-
Learning “O” Helps for Learning More: Handling the Unlabeled Entity Problem for Class-incremental NER, ACL 2023.
-
Teamwork Is Not Always Good: An Empirical Study of Classifier Drift in Class-incremental Information Extraction, ACL (Findings) 2023.
-
ProtoNER: Few Shot Incremental Learning for Named Entity Recognition Using Prototypical Networks, BPM 2023.
-
Task Relation Distillation and Prototypical Pseudo Label for Incremental Named Entity Recognition, CIKM 2023.
-
Continual Named Entity Recognition without Catastrophic Forgetting, CIKM 2023.
-
SKD-NER:Continual Named Entity Recognition via Span-based Knowledge Distillation with Reinforcement Learning, EMNLP 2023.
-
Novel Slot Detection With an Incremental Setting, EMNLP (Findings) 2023.
-
Incremental event detection via an improved knowledge distillation based model, Neurocomputing 2023.
-
Decomposing Logits Distillation for Incremental Named Entity Recognition, SIGIR 2023.
-
Consistent Prototype Learning for Few-Shot Continual Relation Extraction, ACL 2023.
-
Enhancing Continual Relation Extraction via Classifier Decomposition, ACL (Findings) 2023.
-
Improving Continual Relation Extraction by Distinguishing Analogous Semantics, ACL 2023.
-
Serial Contrastive Knowledge Distillation for Continual Few-shot Relation Extraction, ACL (Findings) 2023.
-
ICA-Proto: Iterative Cross Alignment Prototypical Network for Incremental Few-Shot Relation Classification, EACL (Findings) 2023.
-
Serial Contrastive Knowledge Distillation for Continual Few-shot Relation Extraction, ACL (Findings) 2023.
-
Continual Knowledge Distillation for Neural Machine Translation, ACL 2023.
-
Knowledge Transfer in Incremental Learning for Multilingual Neural Machine Translation, ACL 2023.
-
Continual Learning for Multilingual Neural Machine Translation via Dual Importance-based Model Division, EMNLP 2023.
-
Prompts Can Play Lottery Tickets Well: Achieving Lifelong Information Extraction via Lottery Prompt Tuning, ACL 2023.
-
Large-scale Lifelong Learning of In-context Instructions and How to Tackle It, ACL 2023.
-
Generative Replay Inspired by Hippocampal Memory Indexing for Continual Language Learning, EACL 2023.
-
Lifelong Sequence Generation with Dynamic Module Expansion and Adaptation, EMNLP 2023.
-
Orthogonal Subspace Learning for Language Model Continual Learning, EMNLP (Findings) 2023.
-
Sub-network Discovery and Soft-masking for Continual Learning of Mixed Tasks, EMNLP (Findings) 2023.
-
Progressive Prompts: Continual Learning for Language Models, ICLR 2023.
-
Methods for Measuring, Updating, and Visualizing Factual Beliefs in Language Models, EACL 2023.
-
Aging with GRACE: Lifelong Model Editing with Key-Value Adaptors, ICLR 2023.
-
Transformer-Patcher: One Mistake worth One Neuron, ICLR 2023.
-
Mitigating the Alignment Tax of RLHF, Preprint 2023.
-
Tree of Clarifications: Answering Ambiguous Questions with Retrieval-Augmented Large Language Models, EMNLP 2023.
-
Active Retrieval Augmented Generation, EMNLP 2023.
-
Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questions, ACL 2023.
-
Toolalpaca: Generalized tool learning for language models with 3000 simulated cases, Preprint 2023.
-
Gorilla: Large language model connected with massive apis, Preprint 2023.
-
Chameleon: Plug-and-play compositional reasoning with large language models, NIPS 2023.
-
Toolformer: Language Models Can Teach Themselves to Use Tools, NIPS 2023.
-
GPT4Tools: Teaching Large Language Model to Use Tools via Self-instruction, NIPS 2023.
2022
-
Continual Pre-Training Mitigates Forgetting in Language and Vision, Preprint 2022.
-
ELLE: Efficient Lifelong Pre-training for Emerging Data, ACL (Findings) 2022.
-
Lifelong Pretraining: Continually Adapting Language Models to Emerging Corpora, ACL (Workshop) 2022.
-
bert2BERT: Towards Reusable Pretrained Language Models, ACL 2022.
-
Time Waits for No One! Analysis and Challenges of Temporal Misalignment, NAACL 2022.
-
TimeLMs: Diachronic Language Models from Twitter, ACL 2022.
-
Continual Few-shot Intent Detection, COLING 2022.
-
Incremental Intent Detection for Medical Domain with Contrast Replay Networks, ACL (Findings) 2022.
-
Continual training of language models for few-shot learning, EMNLP 2022.
-
Parameter-efficient Continual Learning Framework in Industrial Real-time Text Classification System, NAACL 2022.
-
Prompt Augmented Generative Replay via Supervised Contrastive Learning for Lifelong Intent Detection, NAACL (Findings) 2022.
-
Memory Efficient Continual Learning with Transformers, NIPS 2022.
-
Few-Shot Class-Incremental Learning for Named Entity Recognition, ACL 2022.
-
Learn and Review: Enhancing Continual Named Entity Recognition via Reviewing Synthetic Samples, ACL (Findings) 2022.
-
Incremental Prompting: Episodic Memory Prompt for Lifelong Event Dectection, COLING 2022.
-
Distilling Causal Effect from Miscellaneous Other-Class for Continual Named Entity Recognition, EMNLP 2022.
-
BNU: A Balance-Normalization-Uncertainty Model for Incremental Event Detection, ICASSP 2022.
-
Similarity-Driven Adaptive Prototypical Network for Class-incremental Few-shot Named Entity Recognition, ICTAI 2022.
-
HEFT: A History-Enhanced Feature Transfer framework for incremental event detection, KBS 2022.
-
Consistent Representation Learning for Continual Relation Extraction, ACL (Findings) 2022.
-
Continual Few-shot Relation Learning via Embedding Space Regularization and Data Augmentation, ACL 2022.
-
Less is More: Rethinking State-of-the-art Continual Relation Extraction Models with a Frustratingly Easy but Effective Approach, Preprint 2022.
-
Improving continual relation extraction through prototypical contrastive learning, COLING 2022.
-
Learning Robust Representations for Continual Relation Extraction via Adversarial Class Augmentation, EMNLP 2022.
-
Prompt-based prototypical framework for continual relation extraction, TASLP 2022.
-
Overcoming Catastrophic Forgetting beyond Continual Learning: Balanced Training for Neural Machine Translation, ACL 2022.
-
CLLE: A Benchmark for Continual Language Learning Evaluation in Multilingual Machine Translation, EMNLP (Findings) 2022.
-
Continual Learning of Neural Machine Translation within Low Forgetting Risk Regions, EMNLP 2022.
-
Entropy-Based Vocabulary Substitution for Incremental Learning in Multilingual Neural Machine Translation, EMNLP 2022.
-
Continual Sequence Generation with Adaptive Compositional Modules, ACL 2022.
-
ConTinTin: Continual Learning from Task Instructions, ACL 2022.
-
Continual Prompt Tuning for Dialog State Tracking, ACL 2022.
-
Prompt Conditioned VAE: Enhancing Generative Replay for Lifelong Learning in Task-Oriented Dialogue, EMNLP 2022.
-
Fine-tuned Language Models are Continual Learners, EMNLP 2022.
-
LFPT5: A Unified Framework for Lifelong Few-shot Language Learning Based on Prompt Tuning of T5, ICLR 2022.
-
Plug-and-Play Adaptation for Continuously-updated QA, ACL (Findings) 2022.
2021
-
Temporal Adaptation of BERT and Performance on Downstream Document Classification: Insights from Social Media, EMNLP (Findings) 2021.
-
ECONET: Effective Continual Pretraining of Language Models for Event Temporal Reasoning, EMNLP 2021.
-
Mind the Gap: Assessing Temporal Generalization in Neural Language Models, NIPS 2021.
-
Learning to Solve NLP Tasks in an Incremental Number of Languages, ACL 2021.
-
Adapting BERT for Continual Learning of a Sequence of Aspect Sentiment Classification Tasks, NAACL 2021.
-
CLASSIC: Continual and Contrastive Learning of Aspect Sentiment Classification Tasks, EMNLP 2021.
-
Continual Learning for Text Classification with Information Disentanglement Based Regularization, NAACL 2021.
-
Hyperparameter-free Continuous Learning for Domain Classification in Natural Language Understanding, NAACL 2021.
-
Incremental Few-shot Text Classification with Multi-round New Classes: Formulation, Dataset and System, NAACL 2021.
-
Achieving Forgetting Prevention and Knowledge Transfer in Continual Learning, NIPS 2021.
-
Iterative Network Pruning with Uncertainty Regularization for Lifelong Sentiment Classification, SIGIR 2021.
-
Lifelong Intent Detection via Multi-Strategy Rebalancing, Preprint 2021.
-
Lifelong Knowledge-Enriched Social Event Representation Learning, EACL 2021.
-
Lifelong Learning of Hate Speech Classification on Social Media, NAACL 2021.
-
Continual Learning for Named Entity Recognition, AAAI 2021.
-
Lifelong Event Detection with Knowledge Transfer, EMNLP 2021.
-
Curriculum-meta learning for order-robust continual relation extraction, AAAI 2021.
-
Refining Sample Embeddings with Relation Prototypes to Enhance Continual Relation Extraction, ACL 2021.
-
Continual learning in multilingual nmt via language-specific embeddings, WMT 2021.
-
Continual Learning for Neural Machine Translation, NAACL 2021.
-
Towards Continual Learning for Multilingual Machine Translation via Vocabulary Substitution, NAACL 2021.
-
Rational LAMOL: A rationale-based lifelong learning framework, ACL 2021.
-
Continual Learning for Task-oriented Dialogue System with Iterative Network Pruning, Expanding and Masking, ACL 2021.
-
Learn Continually, Generalize Rapidly: Lifelong Knowledge Accumulation for Few-shot Learning, EMNLP (Findings) 2021.
-
Continual Learning in Task-Oriented Dialogue Systems, EMNLP 2021.
2020
-
Incremental Event Detection via Knowledge Consolidation Networks, EMNLP 2020.
-
Continual Relation Learning via Episodic Memory Activation and Reconsolidation, ACL 2020.
-
Findings of the First Shared Task on Lifelong Learning Machine Translation, WMT 2020.
-
Distill and Replay for Continual Language Learning, COLING 2020.
-
Continual Learning for Natural Language Generation in Task-oriented Dialog Systems, EMNLP (Findings) 2020.
-
LAMOL: LAnguage MOdeling for Lifelong Language Learning, ICLR 2020.
-
Dense Passage Retrieval for Open-Domain Question Answering, EMNLP 2020.
-
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks, NIPS 2020.
2019
-
A Progressive Model to Enable Continual Learning for Semantic Slot Filling, EMNLP 2019.
-
Meta-Learning Improves Lifelong Relation Extraction, RepL4NLP 2019.
-
Sentence Embedding Alignment for Lifelong Relation Extraction, NAACL 2019.
-
From Bilingual to Multilingual Neural Machine Translation by Incremental Training, JASIST 2019.
-
Incremental Learning from Scratch for Task-Oriented Dialogue Systems, ACL 2019.