ACTD: Anchor-Based Cross-Tokenizer Distillation with Residual Regularization

arXiv:2608.29662 · cs.CL · Submitted 2026-08-30 · Read on arXiv

cs.CL

Submitted: 2026-08-30

Updated: 2026-08-30

Code: https://github.com/hyhy892/ACTD

Terminology

• Distillation — 105× in this paper · explained in 2 episodes
• Alignment — 40× in this paper · explained in 1 episode
• Regularization — 24× in this paper · explained in 3 episodes
• Knowledge Distillation — 16× in this paper · explained in 11 episodes
• Benchmark — 15× in this paper · explained in 2 episodes
• Anchor Loss — 9× in this paper · explained in 1 episode
• Anchor Tokens — 7× in this paper · explained in 1 episode
• Robustness — 7× in this paper · explained in 5 episodes
• Anchor-based Cross-Tokenizer Distillation — 5× in this paper
• Anchor Noise — 3× in this paper · explained in 1 episode
• Fine-Tuning — 3× in this paper · explained in 5 episodes
• On-Policy Distillation — 3× in this paper · explained in 3 episodes
• Ablation — 2× in this paper · explained in 1 episode
• Accuracy — 2× in this paper · explained in 1 episode
• Adam — 2× in this paper · explained in 2 episodes
• Generalization — 2× in this paper · explained in 13 episodes
• Tokenization — 2× in this paper · explained in 2 episodes
• Ablation Studies — 1× in this paper · explained in 1 episode
• Ablation study — 1× in this paper · explained in 1 episode
• Adaptation — 1× in this paper · explained in 1 episode
• Consistency — 1× in this paper · explained in 3 episodes
• Core — 1× in this paper · explained in 2 episodes
• Decode — 1× in this paper · explained in 2 episodes
• Distribution Matching — 1× in this paper · explained in 2 episodes
• Generalizability — 1× in this paper · explained in 6 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes
• Optimal Transport — 1× in this paper · explained in 5 episodes
• Over-smoothing — 1× in this paper · explained in 4 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Supervised Fine-Tuning (SFT) — 1× in this paper · explained in 8 episodes
• Transfer Learning — 1× in this paper · explained in 10 episodes
• Transformer — 1× in this paper · explained in 2 episodes
• alpha — 1× in this paper · explained in 1 episode

Sources

• Towards Cross-Tokenizer Distillation: the Universal Logit Distillation Loss for LLMs
• "The Whole Is Greater Than the Sum of Its Parts": A Compatibility-Aware Multi-Teacher CoT Distillation Framework
• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
• ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools
• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
• Distilling the Knowledge in a Neural Network
• Exploring Knowledge Purification in Multi-Teacher Knowledge Distillation for LLMs
• Prefix Teach, Suffix Fade: Local Teachability Collapse in Strong-to-Weak On-Policy Distillation
• Llama 2: Open Foundation and Fine-Tuned Chat Models
• ABKD: Pursuing a Proper Allocation of the Probability Mass in Knowledge Distillation via $\alpha$-$\beta$-Divergence
• Finetuned Language Models Are Zero-Shot Learners
• MiMo-V2-Flash Technical Report
• Qwen3 Technical Report
• VAR-MATH: Probing True Mathematical Reasoning in LLMS via Symbolic Multi-Instance Benchmarks
• Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models

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