Copertina

Disperse-Then-Merge: Pushing the Limits of Instruction Tuning via Alignment Tax Reduction

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Fu, Tingchen, Cai, Deng, Liu, Lemao, Shi, Shuming, Yan, Rui
Natura:	Preprint
Pubblicazione:	2024
Soggetti:	Computation and Language Artificial Intelligence
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

TextBind: Multi-turn Interleaved Multimodal Instruction-following in the Wild
di: Li, Huayang, et al.
Pubblicazione: (2023)

On the Transformations across Reward Model, Parameter Update, and In-Context Prompt
di: Cai, Deng, et al.
Pubblicazione: (2024)

Towards Privacy-Preserving Machine Translation at the Inference Stage: A New Task and Benchmark
di: Shao, Wei, et al.
Pubblicazione: (2026)

From Drafts to Answers: Unlocking LLM Potential via Aggregation Fine-Tuning
di: Li, Yafu, et al.
Pubblicazione: (2025)

Scaling Reasoning, Losing Control: Evaluating Instruction Following in Large Reasoning Models
di: Fu, Tingchen, et al.
Pubblicazione: (2025)

Same Question, Different Words: A Latent Adversarial Framework for Prompt Robustness
di: Fu, Tingchen, et al.
Pubblicazione: (2025)

Selection-p: Self-Supervised Task-Agnostic Prompt Compression for Faithfulness and Transferability
di: Chung, Tsz Ting, et al.
Pubblicazione: (2024)

Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Models
di: Zhang, Yue, et al.
Pubblicazione: (2023)

Pruning via Merging: Compressing LLMs via Manifold Alignment Based Layer Merging
di: Liu, Deyuan, et al.
Pubblicazione: (2024)

A Modular Approach for Clinical SLMs Driven by Synthetic Data with Pre-Instruction Tuning, Model Merging, and Clinical-Tasks Alignment
di: Corbeil, Jean-Philippe, et al.
Pubblicazione: (2025)

Rethinking Table Instruction Tuning
di: Deng, Naihao, et al.
Pubblicazione: (2025)

Merging Triggers, Breaking Backdoors: Defensive Poisoning for Instruction-Tuned Language Models
di: Kim, San, et al.
Pubblicazione: (2026)

Upcycling Instruction Tuning from Dense to Mixture-of-Experts via Parameter Merging
di: Hui, Tingfeng, et al.
Pubblicazione: (2024)

A Closer Look at the Limitations of Instruction Tuning
di: Ghosh, Sreyan, et al.
Pubblicazione: (2024)

OpenSeeker-v2: Pushing the Limits of Search Agents with Informative and High-Difficulty Trajectories
di: Du, Yuwen, et al.
Pubblicazione: (2026)

Human-Instruction-Free LLM Self-Alignment with Limited Samples
di: Guo, Hongyi, et al.
Pubblicazione: (2024)

System-2 Mathematical Reasoning via Enriched Instruction Tuning
di: Cai, Huanqia, et al.
Pubblicazione: (2024)

CrossIn: An Efficient Instruction Tuning Approach for Cross-Lingual Knowledge Alignment
di: Lin, Geyu, et al.
Pubblicazione: (2024)

Instruction Tuning With Loss Over Instructions
di: Shi, Zhengyan, et al.
Pubblicazione: (2024)

Early Detection and Reduction of Memorisation for Domain Adaptation and Instruction Tuning
di: Slack, Dean L., et al.
Pubblicazione: (2025)

SafeMERGE: Preserving Safety Alignment in Fine-Tuned Large Language Models via Selective Layer-Wise Model Merging
di: Djuhera, Aladin, et al.
Pubblicazione: (2025)

LLaVA-PruMerge: Adaptive Token Reduction for Efficient Large Multimodal Models
di: Shang, Yuzhang, et al.
Pubblicazione: (2024)

Pangu Ultra: Pushing the Limits of Dense Large Language Models on Ascend NPUs
di: Yin, Yichun, et al.
Pubblicazione: (2025)

Enhancing Large Language Model for Knowledge Graph Completion via Structure-Aware Alignment-Tuning
di: Liu, Yu, et al.
Pubblicazione: (2025)

Reasoning Pattern Alignment Merging for Adaptive Reasoning
di: Zhong, Zhaofeng, et al.
Pubblicazione: (2026)

Tuning LLMs with Contrastive Alignment Instructions for Machine Translation in Unseen, Low-resource Languages
di: Mao, Zhuoyuan, et al.
Pubblicazione: (2024)

SDA: Steering-Driven Distribution Alignment for Open LLMs without Fine-Tuning
di: Xia, Wei, et al.
Pubblicazione: (2025)

MergeIT: From Selection to Merging for Efficient Instruction Tuning
di: Cai, Hongyi, et al.
Pubblicazione: (2025)

Exploring Format Consistency for Instruction Tuning
di: Liang, Shihao, et al.
Pubblicazione: (2023)

DivLogicEval: A Framework for Benchmarking Logical Reasoning Evaluation in Large Language Models
di: Chung, Tsz Ting, et al.
Pubblicazione: (2025)

XFT: Unlocking the Power of Code Instruction Tuning by Simply Merging Upcycled Mixture-of-Experts
di: Ding, Yifeng, et al.
Pubblicazione: (2024)

InfiniteICL: Breaking the Limit of Context Window Size via Long Short-term Memory Transformation
di: Cao, Bowen, et al.
Pubblicazione: (2025)

Step-On-Feet Tuning: Scaling Self-Alignment of LLMs via Bootstrapping
di: Wang, Haoyu, et al.
Pubblicazione: (2024)

Beyond Distillation: Pushing the Limits of Medical LLM Reasoning with Minimalist Rule-Based RL
di: Liu, Che, et al.
Pubblicazione: (2025)

MegaMath: Pushing the Limits of Open Math Corpora
di: Zhou, Fan, et al.
Pubblicazione: (2025)

BiLLM: Pushing the Limit of Post-Training Quantization for LLMs
di: Huang, Wei, et al.
Pubblicazione: (2024)

The Alignment Tax: Response Homogenization in Aligned LLMs and Its Implications for Uncertainty Estimation
di: Liu, Mingyi
Pubblicazione: (2026)

P-Aligner: Enabling Pre-Alignment of Language Models via Principled Instruction Synthesis
di: Song, Feifan, et al.
Pubblicazione: (2025)

Not All Documents Are What You Need for Extracting Instruction Tuning Data
di: Zhang, Chi, et al.
Pubblicazione: (2025)

Joint Flashback Adaptation for Forgetting-Resistant Instruction Tuning
di: Zhao, Yukun, et al.
Pubblicazione: (2025)