Natural Language Processing Techniques

Últimos artículos

• ModeSwitch-LLM: A Lightweight Phase-Aware Controller for Cross-Mode LLM Inference on a Single GPU

  Aman Sunesh, Ali Alshehhi, Hivansh Dhakne · 25 de mayo de 2026

  ModeSwitch-LLM is a lightweight request-boundary controller for improving single-GPU large language model inference efficiency by routing each request to an appropriate fixed inference mode. Instead of relying on one static serving configuration, the system selects among FP16, quantized modes, specu…

• Instance-Optimal Estimation with Multiple LLM Judges on a Budget

  Junghyun Lee, Sanghwa Kim, Yassir Jedra, Alexandre Prouti\`ere, Se-Young Yun · 25 de mayo de 2026

  Evaluating large language models increasingly relies on LLM-as-a-judge protocols, but such evaluations remain costly: different judges have different prices and reliabilities, and the difficulty of each prompt-response pair can vary substantially. This raises a basic allocation question: under a fix…

• Beyond Words: Multimodal LLM Knows When to Speak

  Zikai Liao, Yi Ouyang, Yi-Lun Lee, Chen-Ping Yu, Yi-Hsuan Tsai, Zhaozheng Yin · 22 de mayo de 2026

  Chatbots via large language models (LLMs) generate fluent responses but often struggle with when to speak, especially for brief, timely listener reactions during ongoing dialogue. We present a multimodal strategy for LLMs, which leverages synchronized video, audio, and text cues to improve conversat…

• Decomposing MXFP4 quantization error for LLM reinforcement learning: reducible bias, recoverable deadzone, and an irreducible floor

  Xiaocan Li, Shiliang Wu, Zheng Shen · 21 de mayo de 2026

  MXFP4 arithmetic can dramatically accelerate reinforcement learning (RL) post-training of large language models (LLMs), yet the quantization error introduces severe accuracy degradation. Existing work treats the quantization error as a monolithic noise term, missing the distinct mechanisms upon inte…

• Efficient numeracy in language models through single-token number embeddings

  Linus Kreitner, Paul Hager, Jonathan Mengedoht, Georgios Kaissis, Daniel Rueckert, Martin J. Menten · 21 de mayo de 2026

  To drive progress in science and engineering, large language models (LLMs) must be able to process large amounts of numerical data and solve long calculations efficiently. This is currently only possible through the use of external tools or extensive reasoning chains, either weakening the numerical …

• How Much Online RL is Enough? Informative Rollouts for Offline Preference Optimization in RLVR

  Richa Verma, Balaraman Ravindran · 21 de mayo de 2026

  Reinforcement Learning from Verifiable Rewards (RLVR) has emerged as a powerful paradigm for reasoning in language models, with GRPO as its primary example. However, GRPO requires continuous online rollout generation, making it computationally expensive and difficult to scale. While Direct Preferenc…

• PlexRL: Cluster-Level Orchestration of Serviceized LLM Execution for RLVR

  Yiqi Zhang, Fangzheng Jiao, Tian Tang, Boyu Tian, Hangyu Wang, Qiaoling Chen, Guoteng Wang, Zhen Jiang, Peng Sun, Ping Zhang, Xiaohe Hu, Ziming Liu, Menghao Zhang, Yanmin Jia, Yang You, Siyuan Feng · 21 de mayo de 2026

  Reinforcement learning with verifiable rewards (RLVR) has recently unlocked strong reasoning capabilities in large language models (LLMs), triggering rapid exploration of new algorithms and data. However, RLVR training is notoriously inefficient: long-tailed rollouts, tool-induced stalls, and asymme…

• Llamas on the Web: Memory-Efficient, Performance-Portable, and Multi-Precision LLM Inference with WebGPU

  Reese Levine, Rithik Sharma, Nikhil Jain, Abhijit Ramesh, Zheyuan Chen, Neha Abbas, James Contini, Tyler Sorensen · 21 de mayo de 2026

  Running language models in the browser presents a unique opportunity to build efficient, private, and portable AI applications, but requires contending with constrained memory availability and heterogeneous hardware targets. To realize this opportunity, we present Llamas on the Web (LlamaWeb), a Web…

• Position: Let's Develop Data Probes to Fundamentally Understand How Data Affects LLM Performance

  Shiqiang Wang, Herbert Woisetschl\"ager, Hans Arno Jacobsen, Mingyue Ji · 21 de mayo de 2026

  Data is fundamental to large language models (LLMs). However, understanding of what makes certain data useful for different stages of an LLM workflow, including training, tuning, alignment, in-context learning, etc., and why, remains an open question. Current approaches rely heavily on extensive exp…

• Learning Abstract World Models with a Group-Structured Latent Space

  Thomas Delliaux, Nguyen-Khanh Vu, Vincent Fran\c{c}ois-Lavet, Elise van der Pol, Emmanuel Rachelson · 20 de mayo de 2026

  Learning meaningful abstract models of Markov Decision Processes (MDPs) is crucial for improving generalization from limited data. In this work, we show how geometric priors can be imposed on the low-dimensional representation manifold of a learned transition model. We incorporate known symmetric st…

• Position: Let's Develop Data Probes to Fundamentally Understand How Data Affects LLM Performance

  Shiqiang Wang, Herbert Woisetschl\"ager, Hans Arno Jacobsen, Mingyue Ji · 20 de mayo de 2026

  Data is fundamental to large language models (LLMs). However, understanding of what makes certain data useful for different stages of an LLM workflow, including training, tuning, alignment, in-context learning, etc., and why, remains an open question. Current approaches rely heavily on extensive exp…

• Borrowed Geometry: Cross-Distribution Head-Importance Fingerprints of Frozen Pretrained Gemma 4 31B

  Abay Bektursun · 20 de mayo de 2026

  Frozen Gemma 4 31B weights pretrained exclusively on text, unmodified, transfer through a thin trainable interface to non-text modalities the substrate has never processed. On the L24--L29 slice (192 attention heads), an English-text TxtCopy attention probe (95 sentences) and per-head ablation impac…

• DynaTrain: Fast Online Parallelism Switching for Elastic LLM Training

  Yuanqing Wang, Yuchen Zhang, Hao Lin, Junhao Hu, Chunyang Zhu, Quanlu Zhang, Boxun Li, Guohao Dai, Zhi Yang, Daning Cheng, Yunquan Zhang, Yu Wang · 20 de mayo de 2026

  Modern large language model (LLM) training is inherently dynamic: resource fluctuations, RLHF phase shifts, and cluster elasticity continually reshape the optimal parallelism layout, posing a significant challenge to existing training frameworks built around a static execution model. We present Dyna…

• Position: Let's Develop Data Probes to Fundamentally Understand How Data Affects LLM Performance

  Shiqiang Wang, Herbert Woisetschl\"ager, Hans Arno Jacobsen, Mingyue Ji · 20 de mayo de 2026

  Data is fundamental to large language models (LLMs). However, understanding of what makes certain data useful for different stages of an LLM workflow, including training, tuning, alignment, in-context learning, etc., and why, remains an open question. Current approaches rely heavily on extensive exp…

• Multi-Token Residual Prediction

  Yufeng Xu, Zishuo Bao, Qian Wang, Zeshen Zhang, Haoqi Zhang, Bowen Peng, Ang Li, Rahul Chalamala, Yucheng Lu · 20 de mayo de 2026

  Diffusion Language Models (DLMs) generate text by iteratively denoising masked token sequences, offering a tradeoff between parallelism and quality compared to autoregressive models. In current practice, the number of tokens decoded per step is controlled by a confidence threshold, and quality degra…

• optimize_anything: A Universal API for Optimizing any Text Parameter

  Lakshya A Agrawal, Donghyun Lee, Shangyin Tan, Wenjie Ma, Karim Elmaaroufi, Rohit Sandadi, Sanjit A. Seshia, Koushik Sen, Dan Klein, Ion Stoica, Joseph E. Gonzalez, Omar Khattab, Alexandros G. Dimakis, Matei Zaharia · 20 de mayo de 2026

  Can a single LLM-based optimization system match specialized tools across fundamentally different domains? We show that when optimization problems are formulated as improving a text artifact evaluated by a scoring function, a single AI-based optimization system-supporting single-task search, multi-t…

• Slot-MLLM: Object-Centric Visual Tokenization for Multimodal LLM

  Donghwan Chi, Hyomin Kim, Yoonjin Oh, Yongjin Kim, Donghoon Lee, Daejin Jo, Jongmin Kim, Junyeob Baek, Sungjin Ahn, Sungwoong Kim · 20 de mayo de 2026

  Recently, multimodal large language models (MLLMs) have emerged as a key approach in achieving artificial general intelligence. In particular, vision-language MLLMs have been developed to generate not only text but also visual outputs from multimodal inputs. This advancement requires efficient image…

• Fine-tuning Large Language Model for Automated Algorithm Design

  Fei Liu, Rui Zhang, Xi Lin, Zhichao Lu, Qingfu Zhang · 20 de mayo de 2026

  The integration of large language models (LLMs) into automated algorithm design has shown promising potential. A prevalent approach embeds LLMs within search routines to iteratively generate and refine candidate algorithms. However, most existing methods rely on off-the-shelf LLMs trained for genera…

• Probabilistic Tiny Recursive Model

  Amin Sghaier, Ali Parviz, Alexia Jolicoeur-Martineau · 20 de mayo de 2026

  Tiny Recursive Models (TRM) solve complex reasoning tasks with a fraction of the parameters of modern large language models (LLMs) by iteratively refining a latent state and final answer. While powerful, their deterministic recursion can lead to convergence at suboptimal solutions, without escape me…

• The Silent Hyperparameter: Quantifying the Impact of Inference Backends on LLM Reproducibility

  David Pape, Jonathan Evertz, Lea Sch\"onherr · 20 de mayo de 2026

  Progress in LLMs is increasingly measured through standardized benchmarks, where state-of-the-art improvements are often separated by fractions of a percentage point. At the same time, the computational cost of evaluating modern LLMs has driven widespread adoption of specialized inference backends, …

• Critique-Guided Distillation for Robust Reasoning via Refinement

  Berkcan Kapusuzoglu, Supriyo Chakraborty, Zain Sarwar, Chia-Hsuan Lee, Sambit Sahu · 20 de mayo de 2026

  Supervised fine-tuning with expert demonstrations often produces models that imitate outputs without internalizing the reasoning processes needed for robust generalization. While critique-based approaches show promise, training models to generate critiques directly, such as Critique Fine-Tuning (CFT…

• GUIDE-VAE: Advancing Data Generation with User Information and Pattern Dictionaries

  Kutay B\"olat, Simon Tindemans · 19 de mayo de 2026

  Generative modelling of multi-user datasets has become prominent in science and engineering. Generating a data point for a given user requires employing user information, and conventional generative models, including variational autoencoders (VAEs), often ignore this. This paper introduces GUIDE-VAE…

• Convex Dataset Valuation for Post-Training

  Siqi Zeng, Christopher Jung, Rui Li, Zhe Kang, Ming Li, Nima Noorshams, Zhigang Wang, Fuchun Peng, Han Zhao, Xue Feng · 19 de mayo de 2026

  Improving LLM performance on downstream tasks sometimes requires leveraging auxiliary datasets during post-training. In practice, however, developers face constraints on compute, labeling, and licensing costs that preclude using all available data, necessitating principled dataset-level selection. T…

• WinQ: Accelerating Quantization-Aware Training of Language Models Around Saddle Points

  Dongyue Li, Zechun Liu, Kai Yi, Zhenshuo Zhang, Changsheng Zhao, Raghuraman Krishnamoorthi, Harshit Khaitan, Hongyang R. Zhang, Steven Li · 19 de mayo de 2026

  Quantization-aware training (QAT) is widely adopted to quantize language models by training full-precision weights using gradients from the quantized model. The main bottleneck is its slow convergence and early performance plateau, particularly below 4-bit-widths. While this problem has been observe…

• RAP: Runtime Adaptive Pruning for LLM Inference

  Huanrong Liu, Chunlin Tian, Xuyang Wei, Qingbiao Li, Li Li · 19 de mayo de 2026

  Large language models (LLMs) excel at language understanding and generation, but their enormous computational and memory requirements hinder deployment. Compression offers a potential solution to mitigate these constraints. However, most existing methods rely on fixed heuristics and thus fail to ada…