Image de couverture de livre

SASFT: Sparse Autoencoder-guided Supervised Finetuning to Mitigate Unexpected Code-Switching in LLMs

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Deng, Boyi, Wan, Yu, Yang, Baosong, Huang, Fei, Wang, Wenjie, Feng, Fuli
Format:	Preprint
Publié:	2025
Sujets:	Computation and Language
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Unveiling Language-Specific Features in Large Language Models via Sparse Autoencoders
par: Deng, Boyi, et autres
Publié: (2025)

Controllable LLM Reasoning via Sparse Autoencoder-Based Steering
par: Fang, Yi, et autres
Publié: (2026)

CrAM: Credibility-Aware Attention Modification in LLMs for Combating Misinformation in RAG
par: Deng, Boyi, et autres
Publié: (2024)

P-MMEval: A Parallel Multilingual Multitask Benchmark for Consistent Evaluation of LLMs
par: Zhang, Yidan, et autres
Publié: (2024)

DLM-Scope: Mechanistic Interpretability of Diffusion Language Models via Sparse Autoencoders
par: Wang, Xu, et autres
Publié: (2026)

Towards Cross-lingual Values Judgment: A Consensus-Pluralism Perspective
par: Chen, Yukun, et autres
Publié: (2026)

Qwen-Scope: Turning Sparse Features into Development Tools for Large Language Models
par: Deng, Boyi, et autres
Publié: (2026)

Counterfactual Debating with Preset Stances for Hallucination Elimination of LLMs
par: Fang, Yi, et autres
Publié: (2024)

Assistant-Guided Mitigation of Teacher Preference Bias in LLM-as-a-Judge
par: Liu, Zhuo, et autres
Publié: (2025)

SAFE: A Sparse Autoencoder-Based Framework for Robust Query Enrichment and Hallucination Mitigation in LLMs
par: Abdaljalil, Samir, et autres
Publié: (2025)

PART: Progressive Alignment Representation Training for Multilingual Speech-To-Text with LLMs
par: Zhang, Pei, et autres
Publié: (2025)

Constrain Alignment with Sparse Autoencoders
par: Yin, Qingyu, et autres
Publié: (2024)

Understanding the Effects of Domain Finetuning on LLMs
par: Tanwar, Eshaan, et autres
Publié: (2025)

Breaking Bad Tokens: Detoxification of LLMs Using Sparse Autoencoders
par: Goyal, Agam, et autres
Publié: (2025)

CultureSynth: A Hierarchical Taxonomy-Guided and Retrieval-Augmented Framework for Cultural Question-Answer Synthesis
par: Zhang, Xinyu, et autres
Publié: (2025)

Improving Task Diversity in Label Efficient Supervised Finetuning of LLMs
par: Arabelly, Abhinav, et autres
Publié: (2025)

Sparse Autoencoders are Capable LLM Jailbreak Mitigators
par: Assogba, Yannick, et autres
Publié: (2026)

Uncovering Cross-Linguistic Disparities in LLMs using Sparse Autoencoders
par: Xuan, Richmond Sin Jing, et autres
Publié: (2025)

Locking Down the Finetuned LLMs Safety
par: Zhu, Minjun, et autres
Publié: (2024)

SCAR: Sparse Conditioned Autoencoders for Concept Detection and Steering in LLMs
par: Härle, Ruben, et autres
Publié: (2024)

Self-Improvement Towards Pareto Optimality: Mitigating Preference Conflicts in Multi-Objective Alignment
par: Li, Moxin, et autres
Publié: (2025)

Investigating and Scaling up Code-Switching for Multilingual Language Model Pre-Training
par: Wang, Zhijun, et autres
Publié: (2025)

Improving Sparse Memory Finetuning
par: Goyal, Satyam, et autres
Publié: (2026)

CultureForest: Understanding and Evaluating Cultural Norm Grounded Reasoning in LLMs
par: Ye, Yangfan, et autres
Publié: (2026)

Mitigating Large Language Model Hallucination with Faithful Finetuning
par: Hu, Minda, et autres
Publié: (2024)

Sparse Autoencoder Insights on Voice Embeddings
par: Pluth, Daniel, et autres
Publié: (2025)

Interpreting and Steering LLMs with Mutual Information-based Explanations on Sparse Autoencoders
par: Wu, Xuansheng, et autres
Publié: (2025)

Right Is Not Enough: The Pitfalls of Outcome Supervision in Training LLMs for Math Reasoning
par: Guo, Jiaxing, et autres
Publié: (2025)

Can Code-Switched Texts Activate a Knowledge Switch in LLMs? A Case Study on English-Korean Code-Switching
par: Kim, Seoyeon, et autres
Publié: (2024)

Randomized Masked Finetuning: An Efficient Way to Mitigate Memorization of PIIs in LLMs
par: Joshi, Kunj, et autres
Publié: (2025)

One Adapts to Any: Meta Reward Modeling for Personalized LLM Alignment
par: Cai, Hongru, et autres
Publié: (2026)

Steering LVLMs via Sparse Autoencoder for Hallucination Mitigation
par: Hua, Zhenglin, et autres
Publié: (2025)

CS3-Bench: Evaluating and Enhancing Speech-to-Speech LLMs for Mandarin-English Code-Switching
par: Liu, Heyang, et autres
Publié: (2025)

Sparse Upcycling: Inference Inefficient Finetuning
par: Doubov, Sasha, et autres
Publié: (2024)

Finetuning LLMs for Comparative Assessment Tasks
par: Raina, Vatsal, et autres
Publié: (2024)

Robo-Instruct: Simulator-Augmented Instruction Alignment For Finetuning Code LLMs
par: Hu, Zichao, et autres
Publié: (2024)

Unveiling Over-Memorization in Finetuning LLMs for Reasoning Tasks
par: Ruan, Zhiwen, et autres
Publié: (2025)

Sampling-Efficient Test-Time Scaling: Self-Estimating the Best-of-N Sampling in Early Decoding
par: Wang, Yiming, et autres
Publié: (2025)

HellaSwag-Pro: A Large-Scale Bilingual Benchmark for Evaluating the Robustness of LLMs in Commonsense Reasoning
par: Li, Xiaoyuan, et autres
Publié: (2025)

EAVE: Efficient Product Attribute Value Extraction via Lightweight Sparse-layer Interaction
par: Yang, Li, et autres
Publié: (2024)