Estimating and Orthogonalizing Unknown Pre-training Gradients for Continual Fine-tuning of Large Language Models

arXiv:2609.30935 · cs.CL, cs.AI, cs.LG · Submitted 2026-09-25 · Read on arXiv

cs.CL, cs.AI, cs.LG

Submitted: 2026-09-25

Updated: 2026-09-25

Comments: Accepted by NeurIPS 2026. 29 pages, 3 figures. Code: https://github.com/wangbing1416/EoupCT

Code: https://github.com/wangbing1416/EoupCT

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Fine-Tuning — 57× in this paper · explained in 5 episodes
• Catastrophic Forgetting — 26× in this paper · explained in 9 episodes
• Adaptation — 14× in this paper · explained in 1 episode
• Benchmark — 13× in this paper · explained in 2 episodes
• Continual Learning — 13× in this paper · explained in 3 episodes
• Accuracy — 12× in this paper · explained in 1 episode
• Multi-objective Optimization — 11× in this paper · explained in 5 episodes
• Core — 7× in this paper · explained in 2 episodes
• Distillation — 7× in this paper · explained in 2 episodes
• Ablation — 5× in this paper · explained in 1 episode
• Downstream Tasks — 5× in this paper · explained in 2 episodes
• Generalization — 5× in this paper · explained in 13 episodes
• Alignment — 4× in this paper
• Computational overhead — 4× in this paper · explained in 2 episodes
• Embeddings — 4× in this paper · explained in 2 episodes
• Knowledge Distillation — 4× in this paper · explained in 15 episodes
• Regularization — 4× in this paper · explained in 3 episodes
• Instruction Tuning — 3× in this paper · explained in 3 episodes
• LoRA Fine-Tuning — 3× in this paper · explained in 3 episodes
• Adaptability — 2× in this paper · explained in 1 episode
• Generative Model — 2× in this paper · explained in 2 episodes
• Institutional Review Board — 2× in this paper
• Large Language Models (LLMs) — 2× in this paper · explained in 8 episodes
• Loss Function — 2× in this paper · explained in 2 episodes
• Transformer — 2× in this paper · explained in 2 episodes
• Ablation Studies — 1× in this paper · explained in 1 episode
• Adversarial Training — 1× in this paper
• Attention Mechanism — 1× in this paper · explained in 3 episodes
• Cosine Similarity — 1× in this paper · explained in 3 episodes
• Decoupling — 1× in this paper · explained in 2 episodes
• Domain Adaptation — 1× in this paper · explained in 5 episodes
• Generalization Bounds — 1× in this paper · explained in 2 episodes
• Generative Models — 1× in this paper · explained in 3 episodes
• Gradient Flow — 1× in this paper · explained in 2 episodes
• Latent Representation — 1× in this paper · explained in 2 episodes
• Latent Representations — 1× in this paper · explained in 3 episodes
• Local Optimization — 1× in this paper · explained in 2 episodes
• Low-Rank Adaptation (LoRA) — 1× in this paper · explained in 16 episodes
• Objective Function — 1× in this paper · explained in 2 episodes
• On-Policy Distillation — 1× in this paper · explained in 3 episodes
• Parameter-Efficient Fine-Tuning — 1× in this paper · explained in 3 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Transformer Architecture — 1× in this paper · explained in 3 episodes

Sources

• An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning
• Gemma 2: Improving Open Language Models at a Practical Size
• Qwen3 Technical Report
• Distribution-Aligned Sequence Distillation for Superior Long-CoT Reasoning

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering