Model merging dramatically reduces storage and computational resources by combining multiple expert models into a single multi-task model. Although recent model merging methods have shown promising results, they struggle to maintain performance gains as the number of merged models increases. In this paper, we investigate the key obstacles that limit the scalability of model merging when integrating a large number of expert models. First, we prove that there is an upper bound on model merging. Further theoretical analysis reveals that the limited effective parameter space imposes a strict constraint on the number of models that can be successfully merged. Gaussian Width shows that the marginal benefit of merging additional models diminishes according to a strictly concave function. This implies that the effective parameter space becomes rapidly saturated as the number of merged models increases. Furthermore, using Approximate Kinematics Theory, we prove the existence of a unique optimal threshold beyond which adding more models does not yield significant performance improvements. At the same time, we introduce a straightforward Reparameterized Heavy-Tailed method (RHT) to extend the coverage of the merged model, thereby enhancing its performance. Empirical results on 12 benchmarks, including both knowledge-intensive and general-purpose tasks, validate our theoretical analysis. We believe that these results spark further research beyond the current scope of model merging. The source code is in the Github repository: https://github.com/wzj1718/ModelMergingAnalysis.

Subject-to-Video (S2V) generation aims to create videos that faithfully incorporate reference content, providing enhanced flexibility in the production of videos. To establish the infrastructure for S2V generation, we propose OpenS2V-Nexus, consisting of (i) OpenS2V-Eval, a fine-grained benchmark, and (ii) OpenS2V-5M, a million-scale dataset. In contrast to existing S2V benchmarks inherited from VBench that focus on global and coarse-grained assessment of generated videos, OpenS2V-Eval focuses on the model's ability to generate subject-consistent videos with natural subject appearance and identity fidelity. For these purposes, OpenS2V-Eval introduces 180 prompts from seven major categories of S2V, which incorporate both real and synthetic test data. Furthermore, to accurately align human preferences with S2V benchmarks, we propose three automatic metrics, NexusScore, NaturalScore and GmeScore, to separately quantify subject consistency, naturalness, and text relevance in generated videos. Building on this, we conduct a comprehensive evaluation of 18 representative S2V models, highlighting their strengths and weaknesses across different content. Moreover, we create the first open-source large-scale S2V generation dataset OpenS2V-5M, which consists of five million high-quality 720P subject-text-video triples. Specifically, we ensure subject-information diversity in our dataset by (1) segmenting subjects and building pairing information via cross-video associations and (2) prompting GPT-Image-1 on raw frames to synthesize multi-view representations. Through OpenS2V-Nexus, we deliver a robust infrastructure to accelerate future S2V generation research.

We investigate how to teach large language models (LLMs) to perform scientific reasoning by leveraging expert discussions as a learning signal. Focusing on the genomics domain, we develop an automated pipeline to extract trainable data and introduce Genome-Bench, a new benchmark constructed from over a decade of scientific forum discussions on genome engineering. Our pipeline transforms raw interactions into a reinforcement learning-friendly multiple-choice questions format, supported by 3000+ high-quality question-answer pairs spanning foundational biology, experimental troubleshooting, tool usage, and beyond. We fine-tune an LLM using RL with a rule-based reward signal derived from the synthetic MCQ dataset to enhance domain-specific reasoning. Our results show that reinforcement learning from scientific discussions improves model performance by over 15% compared to the base model on Genome-Bench, narrowing the gap between open-source LLMs and expert-level reasoning. To our knowledge, this is the first end-to-end pipeline for teaching LLMs to reason from scientific discussions, with promising potential for generalization across scientific domains beyond biology.

Large language models (LLMs) represent a new paradigm for processing unstructured data, with applications across an unprecedented range of domains. In this paper, we address, through two arguments, whether the development and application of LLMs would genuinely benefit from foundational contributions from the statistics discipline. First, we argue affirmatively, beginning with the observation that LLMs are inherently statistical models due to their profound data dependency and stochastic generation processes, where statistical insights are naturally essential for handling variability and uncertainty. Second, we argue that the persistent black-box nature of LLMs -- stemming from their immense scale, architectural complexity, and development practices often prioritizing empirical performance over theoretical interpretability -- renders closed-form or purely mechanistic analyses generally intractable, thereby necessitating statistical approaches due to their flexibility and often demonstrated effectiveness. To substantiate these arguments, the paper outlines several research areas -- including alignment, watermarking, uncertainty quantification, evaluation, and data mixture optimization -- where statistical methodologies are critically needed and are already beginning to make valuable contributions. We conclude with a discussion suggesting that statistical research concerning LLMs will likely form a diverse ``mosaic'' of specialized topics rather than deriving from a single unifying theory, and highlighting the importance of timely engagement by our statistics community in LLM research.

High-quality long-context instruction data is essential for aligning long-context large language models (LLMs). Despite the public release of models like Qwen and Llama, their long-context instruction data remains proprietary. Human annotation is costly and challenging, while template-based synthesis methods limit scale, diversity, and quality. We introduce LongMagpie, a self-synthesis framework that automatically generates large-scale long-context instruction data. Our key insight is that aligned long-context LLMs, when presented with a document followed by special tokens preceding a user turn, auto-regressively generate contextually relevant queries. By harvesting these document-query pairs and the model's responses, LongMagpie produces high-quality instructions without human effort. Experiments on HELMET, RULER, and Longbench v2 demonstrate that LongMagpie achieves leading performance on long-context tasks while maintaining competitive performance on short-context tasks, establishing it as a simple and effective approach for open, diverse, and scalable long-context instruction data synthesis.

Playing video games requires perception, memory, and planning, exactly the faculties modern large language model (LLM) agents are expected to master. We study the major challenges in using popular video games to evaluate modern LLMs and find that directly dropping LLMs into games cannot make an effective evaluation, for three reasons -- brittle vision perception, prompt sensitivity, and potential data contamination. We introduce lmgame-Bench to turn games into reliable evaluations. lmgame-Bench features a suite of platformer, puzzle, and narrative games delivered through a unified Gym-style API and paired with lightweight perception and memory scaffolds, and is designed to stabilize prompt variance and remove contamination. Across 13 leading models, we show lmgame-Bench is challenging while still separating models well. Correlation analysis shows that every game probes a unique blend of capabilities often tested in isolation elsewhere. More interestingly, performing reinforcement learning on a single game from lmgame-Bench transfers both to unseen games and to external planning tasks. Our evaluation code is available at https://github.com/lmgame-org/GamingAgent/lmgame-bench.

Large Language Models (LLMs) demonstrate impressive capabilities but lack robust temporal intelligence, struggling to integrate reasoning about the past with predictions and plausible generations of the future. Meanwhile, existing methods typically target isolated temporal skills, such as question answering about past events or basic forecasting, and exhibit poor generalization, particularly when dealing with events beyond their knowledge cutoff or requiring creative foresight. To address these limitations, we introduce \textit{Time-R1}, the first framework to endow a moderate-sized (3B-parameter) LLM with comprehensive temporal abilities: understanding, prediction, and creative generation. Our approach features a novel three-stage development path; the first two constitute a \textit{reinforcement learning (RL) curriculum} driven by a meticulously designed dynamic rule-based reward system. This framework progressively builds (1) foundational temporal understanding and logical event-time mappings from historical data, (2) future event prediction skills for events beyond its knowledge cutoff, and finally (3) enables remarkable generalization to creative future scenario generation without any fine-tuning. Strikingly, experiments demonstrate that Time-R1 outperforms models over 200 times larger, including the state-of-the-art 671B DeepSeek-R1, on highly challenging future event prediction and creative scenario generation benchmarks. This work provides strong evidence that thoughtfully engineered, progressive RL fine-tuning allows smaller, efficient models to achieve superior temporal performance, offering a practical and scalable path towards truly time-aware AI. To foster further research, we also release \textit{Time-Bench}, a large-scale multi-task temporal reasoning dataset derived from 10 years of news data, and our series of \textit{Time-R1} checkpoints.

Current speech-LLMs exhibit limited capability in contextual reasoning alongside paralinguistic understanding, primarily due to the lack of Question-Answer (QA) datasets that cover both aspects. We propose a novel framework for dataset generation from in-the-wild speech data, that integrates contextual reasoning with paralinguistic information. It consists of a pseudo par-alinguistic label-based data condensation of in-the-wild speech and LLM-based Contextual Paralinguistic QA (CPQA) generation. The effectiveness is validated by a strong correlation in evaluations of the Qwen2-Audio-7B-Instruct model on a dataset created by our framework and human-generated CPQA dataset. The results also reveal the speech-LLM's limitations in handling empathetic reasoning tasks, highlighting the need for such datasets and more robust models. The proposed framework is first of its kind and has potential in training more robust speech-LLMs with paralinguistic reasoning capabilities.

Each year, tens of millions of essays are written and graded in college-level English courses. Students are asked to analyze literary and cultural texts through a process known as close reading, in which they gather textual details to formulate evidence-based arguments. Despite being viewed as a basis for critical thinking and widely adopted as a required element of university coursework, close reading has never been evaluated on large language models (LLMs), and multi-discipline benchmarks like MMLU do not include literature as a subject. To fill this gap, we present KRISTEVA, the first close reading benchmark for evaluating interpretive reasoning, consisting of 1331 multiple-choice questions adapted from classroom data. With KRISTEVA, we propose three progressively more difficult sets of tasks to approximate different elements of the close reading process, which we use to test how well LLMs may seem to understand and reason about literary works: 1) extracting stylistic features, 2) retrieving relevant contextual information from parametric knowledge, and 3) multi-hop reasoning between style and external contexts. Our baseline results find that, while state-of-the-art LLMs possess some college-level close reading competency (accuracy 49.7% - 69.7%), their performances still trail those of experienced human evaluators on 10 out of our 11 tasks.

For this week's Open Questions column, Cal Newport is filling in for Joshua Rothman. In the spring of 1940, Isaac Asimov, who had just turned twenty, published a short story titled "Strange Playfellow." It was about an artificially intelligent machine named Robbie that acts as a companion for Gloria, a young girl. Asimov was not the first to explore such technology. In Karel Čapek's play "R.U.R.," which débuted in 1921 and introduced the term "robot," artificial men overthrow humanity, and in Edmond Hamilton's 1926 short story "The Metal Giants" machines heartlessly smash buildings to rubble.