English
Related papers

Related papers: Where does output diversity collapse in post-train…

200 papers

Reinforcement learning has been central to recent advances in large language model reasoning, but most algorithms rely on on-policy training that demands fresh rollouts at every update, limiting efficiency and scalability. Asynchronous RL…

Machine Learning · Computer Science 2025-10-29 Haizhong Zheng , Jiawei Zhao , Beidi Chen

Pre-trained language models (LMs) are capable of in-context learning (ICL): they can adapt to a task with only a few examples given in the prompt without any parameter update. However, it is unclear where this capability comes from as there…

Computation and Language · Computer Science 2024-02-21 Yanda Chen , Chen Zhao , Zhou Yu , Kathleen McKeown , He He

Given the ease of creating synthetic data from machine learning models, new models can be potentially trained on synthetic data generated by previous models. This recursive training process raises concerns about the long-term impact on…

Machine Learning · Computer Science 2024-12-24 Ananda Theertha Suresh , Andrew Thangaraj , Aditya Nanda Kishore Khandavally

Diversity in training data, architecture, and providers is assumed to mitigate homogeneity in LLMs. However, we lack empirical evidence on whether different LLMs differ meaningfully. We conduct a large-scale empirical evaluation on over 350…

Computation and Language · Computer Science 2025-06-10 Elliot Kim , Avi Garg , Kenny Peng , Nikhil Garg

Group Relative Policy Optimization (GRPO) trains Chain-of-Thought reasoning with verifiable rewards, but estimating thought-level advantages without value functions often suffers from high variance. Although tree-style branching is used in…

Computation and Language · Computer Science 2026-02-06 Hongcheng Wang , Yinuo Huang , Sukai Wang , Guanghui Ren , Hao Dong

Large Language Models (LLMs) that undergo recursive training on synthetically generated data are susceptible to model collapse, a phenomenon marked by the generation of meaningless output. Existing research has examined this issue from…

Computation and Language · Computer Science 2026-03-17 Konstantinos F. Xylogiannopoulos , Petros Xanthopoulos , Panagiotis Karampelas , Georgios A. Bakamitsos

Continual learning for large language models is typically evaluated through accuracy retention under sequential fine-tuning. We argue that this perspective is incomplete, because uncertainty reliability can degrade earlier and more sharply…

Machine Learning · Computer Science 2026-04-28 Ibne Farabi Shihab , Sanjeda Akter , Anuj Sharma

Supervised fine-tuning (SFT) on chain-of-thought data is an essential post-training step for reasoning language models. Standard machine learning intuition suggests that training with more unique training samples yields better…

Computation and Language · Computer Science 2026-02-12 Dawid J. Kopiczko , Sagar Vaze , Tijmen Blankevoort , Yuki M. Asano

Reinforcement Learning (RL)-based post-training has significantly advanced the complex reasoning capabilities of language models, fostering sophisticated self-reflection processes. However, this ``slow thinking'' paradigm presents a…

Machine Learning · Computer Science 2025-06-24 Xu Wan , Wei Wang , Wenyue Xu , Wotao Yin , Jie Song , Mingyang Sun

Continual instruction tuning(CIT) during the post-training phase is crucial for adapting multimodal large language models (MLLMs) to evolving real-world demands. However, the progress is hampered by the lack of benchmarks with rigorous,…

Computation and Language · Computer Science 2026-02-16 Haiyun Guo , Zhiyan Hou , Yandu Sun , Jinghan He , Yu Chen , Yuzhe Zhou , Yuheng Jia , Jinqiao Wang , Tat-Seng Chua

Large language models (LLMs) tend to generate homogenous texts, which may impact the diversity of knowledge generated across different outputs. Given their potential to replace existing forms of knowledge acquisition, this poses a risk of…

Chain-of-thought (CoT) prompting has become central to mathematical reasoning in large language models, yet models remain brittle to early errors: a single arithmetic slip or unjustified inference typically propagates uncorrected to an…

Machine Learning · Computer Science 2025-12-22 Saraswathy Amjith , Mihika Dusad , Neha Muramalla , Shweta Shah

Intermediate token generation (ITG), where a model produces output before the solution, has been proposed as a method to improve the performance of language models on reasoning tasks. While these reasoning traces or Chain of Thoughts (CoTs)…

Artificial Intelligence · Computer Science 2025-09-10 Vardhan Palod , Karthik Valmeekam , Kaya Stechly , Subbarao Kambhampati

Modern reasoning language models generate dense, sequential chain-of-thought traces implicitly assuming that every token contributes and that steps must be consumed in order. We challenge both assumptions through a systematic intervention…

Computation and Language · Computer Science 2026-05-11 Yi-Chang Chen , Feng-Ting Liao , Da-shan Shiu , Hung-yi Lee

Meta-learning aims to leverage information across related tasks to improve prediction on unlabeled data for new tasks when only a small number of labeled observations are available ("few-shot" learning). Increased task diversity is often…

Machine Learning · Statistics 2026-01-16 Saptati Datta , Nicolas W. Hengartner , Yulia Pimonova , Natalie E. Klein , Nicholas Lubbers

Recent success of pre-trained language models (LMs) has spurred widespread interest in the language capabilities that they possess. However, efforts to understand whether LM representations are useful for symbolic reasoning tasks have been…

Computation and Language · Computer Science 2020-11-20 Alon Talmor , Yanai Elazar , Yoav Goldberg , Jonathan Berant

Reinforcement learning (RL) has shown extraordinary potential in aligning diffusion models to downstream tasks, yet most of them still suffer from significant reward hacking, which degrades generative diversity and quality by inducing…

Machine Learning · Computer Science 2026-05-14 Jiaming Li , Chenyu Zhu , Nanxi Yi , Youjun Bao , Li Sun , Quanying Lv , Xiang Fang , Daizong Liu , Jianjun Li , Kun He , Bowen Zhou , Zhiyuan Ma

We aim to develop a fundamental understanding of modality collapse, a recently observed empirical phenomenon wherein models trained for multimodal fusion tend to rely only on a subset of the modalities, ignoring the rest. We show that…

Machine Learning · Computer Science 2025-08-18 Abhra Chaudhuri , Anjan Dutta , Tu Bui , Serban Georgescu

The remarkable performance of models like the OpenAI o1 can be attributed to their ability to emulate human-like long-time thinking during inference. These models employ extended chain-of-thought (CoT) processes, exploring multiple…

Computation and Language · Computer Science 2025-02-04 Xingyu Chen , Jiahao Xu , Tian Liang , Zhiwei He , Jianhui Pang , Dian Yu , Linfeng Song , Qiuzhi Liu , Mengfei Zhou , Zhuosheng Zhang , Rui Wang , Zhaopeng Tu , Haitao Mi , Dong Yu

Multilingual Large Language Models considerably changed how technologies influence language. While previous technologies could mediate or assist humans, there is now a tendency to offload the task of writing itself to these technologies,…

Computation and Language · Computer Science 2026-04-24 Eva Vanmassenhove