Copertina

Extracting Small Translation Specialists from LLMs by Aggressively Pruning Experts

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Martin, Liu O., Bandarkar, Lucas, Peng, Nanyun
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Computation and Language Artificial Intelligence Machine Learning I.2.7
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

The Unreasonable Effectiveness of Model Merging for Cross-Lingual Transfer in LLMs
di: Bandarkar, Lucas, et al.
Pubblicazione: (2025)

The Belebele Benchmark: a Parallel Reading Comprehension Dataset in 122 Language Variants
di: Bandarkar, Lucas, et al.
Pubblicazione: (2023)

The Expert Strikes Back: Interpreting Mixture-of-Experts Language Models at Expert Level
di: Herbst, Jeremy, et al.
Pubblicazione: (2026)

Evaluating the Efficacy of Hybrid Deep Learning Models in Distinguishing AI-Generated Text
di: Oketunji, Abiodun Finbarrs
Pubblicazione: (2023)

Slim-SC: Thought Pruning for Efficient Scaling with Self-Consistency
di: Hong, Colin, et al.
Pubblicazione: (2025)

SODA: A Natural Language Processing Package to Extract Social Determinants of Health for Cancer Studies
di: Yu, Zehao, et al.
Pubblicazione: (2022)

How Many Parameters Does Your Task Really Need? Task Specific Pruning with LLM-Sieve
di: Reda, Waleed, et al.
Pubblicazione: (2025)

Every Sample Matters: Leveraging Mixture-of-Experts and High-Quality Data for Efficient and Accurate Code LLM
di: Codefuse, et al.
Pubblicazione: (2025)

Path-Lock Expert: Separating Reasoning Mode in Hybrid Thinking via Architecture-Level Separation
di: Wang, Shouren, et al.
Pubblicazione: (2026)

The Readout Shortcut: Positional Number Copying Dominates Arithmetic CoT Readout in Small Language Models
di: Liu, Ming
Pubblicazione: (2026)

Large Language Model (LLM) Bias Index -- LLMBI
di: Oketunji, Abiodun Finbarrs, et al.
Pubblicazione: (2023)

Consistency Evaluation of News Article Summaries Generated by Large (and Small) Language Models
di: Gilhuly, Colleen, et al.
Pubblicazione: (2025)

The Metacognitive Probe: Five Behavioural Calibration Diagnostics for LLMs
di: Oliveira, Rafael C. T.
Pubblicazione: (2026)

Evaluating Long Range Dependency Handling in Code Generation LLMs
di: Assogba, Yannick, et al.
Pubblicazione: (2024)

Positional Failures in Long-Context LLMs: A Blind Spot in Reasoning Benchmarks
di: Zhang, Chuyifei, et al.
Pubblicazione: (2026)

AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders
di: Wu, Zhengxuan, et al.
Pubblicazione: (2025)

IdentifyMe: A Challenging Long-Context Mention Resolution Benchmark for LLMs
di: Manikantan, Kawshik, et al.
Pubblicazione: (2024)

RomanLens: The Role Of Latent Romanization In Multilinguality In LLMs
di: Saji, Alan, et al.
Pubblicazione: (2025)

REAP the Experts: Why Pruning Prevails for One-Shot MoE compression
di: Lasby, Mike, et al.
Pubblicazione: (2025)

ALBA: A European Portuguese Benchmark for Evaluating Language and Linguistic Dimensions in Generative LLMs
di: Vieira, Inês, et al.
Pubblicazione: (2026)

GMoE: Empowering LLMs Fine-Tuning via MoE Graph Collaboration
di: Bai, Ting, et al.
Pubblicazione: (2024)

Multilingual LLMs Inherently Reward In-Language Time-Sensitive Semantic Alignment for Low-Resource Languages
di: Bajpai, Ashutosh, et al.
Pubblicazione: (2024)

Contextual Integrity in LLMs via Reasoning and Reinforcement Learning
di: Lan, Guangchen, et al.
Pubblicazione: (2025)

Improving Influence-based Instruction Tuning Data Selection for Balanced Learning of Diverse Capabilities
di: Dai, Qirun, et al.
Pubblicazione: (2025)

Pressure-Testing Deception Probes in LLMs: Scaling, Robustness, and the Geometry of Deceptive Representations
di: Kumar, Sachin
Pubblicazione: (2026)

CopySpec: Accelerating LLMs with Speculative Copy-and-Paste Without Compromising Quality
di: Dumitru, Razvan-Gabriel, et al.
Pubblicazione: (2025)

Adapting While Learning: Grounding LLMs for Scientific Problems with Intelligent Tool Usage Adaptation
di: Lyu, Bohan, et al.
Pubblicazione: (2024)

R2R: Efficiently Navigating Divergent Reasoning Paths with Small-Large Model Token Routing
di: Fu, Tianyu, et al.
Pubblicazione: (2025)

Control Reinforcement Learning: Interpretable Token-Level Steering of LLMs via Sparse Autoencoder Features
di: Cho, Seonglae, et al.
Pubblicazione: (2026)

Fine-Tuned 'Small' LLMs (Still) Significantly Outperform Zero-Shot Generative AI Models in Text Classification
di: Bucher, Martin Juan José, et al.
Pubblicazione: (2024)

Layer-Wise Quantization: A Pragmatic and Effective Method for Quantizing LLMs Beyond Integer Bit-Levels
di: Dumitru, Razvan-Gabriel, et al.
Pubblicazione: (2024)

Danoliteracy of Generative Large Language Models
di: Holm, Søren Vejlgaard, et al.
Pubblicazione: (2024)

Large Language Models Generate Harmful Content Using a Distinct, Unified Mechanism
di: Orgad, Hadas, et al.
Pubblicazione: (2026)

CodingTeachLLM: Empowering LLM's Coding Ability via AST Prior Knowledge
di: Chen, Zhangquan, et al.
Pubblicazione: (2024)

Entropy-Based Measurement of Value Drift and Alignment Work in Large Language Models
di: Fadli, Samih
Pubblicazione: (2025)

Self-Supervised Position Debiasing for Large Language Models
di: Liu, Zhongkun, et al.
Pubblicazione: (2024)

Does LLM Alignment Really Need Diversity? An Empirical Study of Adapting RLVR Methods for Moral Reasoning
di: Zhang, Zhaowei, et al.
Pubblicazione: (2026)

Generalizing Test-time Compute-optimal Scaling as an Optimizable Graph
di: Wang, Fali, et al.
Pubblicazione: (2025)

AgentTTS: Large Language Model Agent for Test-time Compute-optimal Scaling Strategy in Complex Tasks
di: Wang, Fali, et al.
Pubblicazione: (2025)

ProSparse: Introducing and Enhancing Intrinsic Activation Sparsity within Large Language Models
di: Song, Chenyang, et al.
Pubblicazione: (2024)