GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models

We present GLM-4.5, an open-source Mixture-of-Experts (MoE) large language model with 355B total parameters and 32B activated parameters, featuring a hybrid reasoning method that supports both thinking and direct response modes. Through multi-stage training on 23T tokens and comprehensive post-training with expert model iteration and reinforcement learning, GLM-4.5 achieves strong performance across agentic, reasoning, and coding (ARC) tasks, scoring 70.1% on TAU-Bench, 91.0% on AIME 24, and 64.2% on SWE-bench Verified. With much fewer parameters than several competitors, GLM-4.5 ranks 3rd overall among all evaluated models and 2nd on agentic benchmarks. We release both GLM-4.5 (355B parameters) and a compact version, GLM-4.5-Air (106B parameters), to advance research in reasoning and agentic AI systems. Code, models, and more information are available at https://github.com/zai-org/GLM-4.5.

MV-Debate: Multi-view Agent Debate with Dynamic Reflection Gating for Multimodal Harmful Content Detection in Social Media

Social media has evolved into a complex multimodal environment where text, images, and other signals interact to shape nuanced meanings, often concealing harmful intent. Identifying such intent, whether sarcasm, hate speech, or misinformation, remains challenging due to cross-modal contradictions, rapid cultural shifts, and subtle pragmatic cues. To address these challenges, we propose MV-Debate, a multi-view agent debate framework with dynamic reflection gating for unified multimodal harmful content detection. MV-Debate assembles four complementary debate agents, a surface analyst, a deep reasoner, a modality contrast, and a social contextualist, to analyze content from diverse interpretive perspectives. Through iterative debate and reflection, the agents refine responses under a reflection-gain criterion, ensuring both accuracy and efficiency. Experiments on three benchmark datasets demonstrate that MV-Debate significantly outperforms strong single-model and existing multi-agent debate baselines. This work highlights the promise of multi-agent debate in advancing reliable social intent detection in safety-critical online contexts.

TreeRL: LLM Reinforcement Learning with On-Policy Tree Search

Reinforcement learning (RL) with tree search has demonstrated superior performance in traditional reasoning tasks. Compared to conventional independent chain sampling strategies with outcome supervision, tree search enables better exploration of the reasoning space and provides dense, on-policy process rewards during RL training but remains under-explored in On-Policy LLM RL. We propose TreeRL, a reinforcement learning framework that directly incorporates on-policy tree search for RL training. Our approach includes intermediate supervision and eliminates the need for a separate reward model training. Existing approaches typically train a separate process reward model, which can suffer from distribution mismatch and reward hacking. We also introduce a cost-effective tree search approach that achieves higher search efficiency under the same generation token budget by strategically branching from high-uncertainty intermediate steps rather than using random branching. Experiments on challenging math and code reasoning benchmarks demonstrate that TreeRL achieves superior performance compared to traditional ChainRL, highlighting the potential of tree search for LLM. TreeRL is open-sourced at https://github.com/THUDM/TreeRL.

Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?

Reinforcement Learning with Verifiable Rewards (RLVR) has recently demonstrated notable success in enhancing the reasoning capabilities of LLMs, particularly in mathematics and programming tasks. It is widely believed that RLVR enables LLMs to continuously self-improve, thus acquiring novel reasoning abilities that exceed corresponding base models' capacity. In this study, however, we critically re-examines this assumption by measuring the pass@\textit{k} metric with large values of \textit{k} to explore the reasoning capability boundary of the models across a wide range of model families and benchmarks. Surprisingly, the RL does \emph{not}, in fact, elicit fundamentally new reasoning patterns. While RL-trained models outperform their base models at smaller values of $k$ (\eg, $k$=1), base models can achieve a comparable or even higher pass@$k$ score compared to their RL counterparts at large $k$ values. The reasoning paths generated by RL-trained models are already included in the base models' sampling distribution, suggesting that most reasoning abilities manifested in RL-trained models are already obtained by base models. Further analysis shows that RL training boosts the performance by biasing the model's output distribution toward paths that are more likely to yield rewards, therefore sampling correct responses more efficiently. But this also results in a narrower reasoning capability boundary compared to base models. Similar results are observed in visual reasoning tasks trained with RLVR. Moreover, we find that distillation can genuinely introduce new knowledge into the model, different from RLVR. These findings underscore a critical limitation of RLVR in advancing LLM reasoning abilities which requires us to fundamentally rethink the impact of RL training in reasoning LLMs and the need of a better paradigm. Project Page: https://limit-of-RLVR.github.io

Towards Understanding Text Hallucination of Diffusion Models via Local Generation Bias

Score-based diffusion models have achieved incredible performance in generating realistic images, audio, and video data. While these models produce high-quality samples with impressive details, they often introduce unrealistic artifacts, such as distorted fingers or hallucinated texts with no meaning. This paper focuses on textual hallucinations, where diffusion models correctly generate individual symbols but assemble them in a nonsensical manner. Through experimental probing, we consistently observe that such phenomenon is attributed it to the network's local generation bias. Denoising networks tend to produce outputs that rely heavily on highly correlated local regions, particularly when different dimensions of the data distribution are nearly pairwise independent. This behavior leads to a generation process that decomposes the global distribution into separate, independent distributions for each symbol, ultimately failing to capture the global structure, including underlying grammar. Intriguingly, this bias persists across various denoising network architectures including MLP and transformers which have the structure to model global dependency. These findings also provide insights into understanding other types of hallucinations, extending beyond text, as a result of implicit biases in the denoising models. Additionally, we theoretically analyze the training dynamics for a specific case involving a two-layer MLP learning parity points on a hypercube, offering an explanation of its underlying mechanism.

VVRec: Reconstruction Attacks on DL-based Volumetric Video Upstreaming via Latent Diffusion Model with Gamma Distribution

With the popularity of 3D volumetric video applications, such as Autonomous Driving, Virtual Reality, and Mixed Reality, current developers have turned to deep learning for compressing volumetric video frames, i.e., point clouds for video upstreaming. The latest deep learning-based solutions offer higher efficiency, lower distortion, and better hardware support compared to traditional ones like MPEG and JPEG. However, privacy threats arise, especially reconstruction attacks targeting to recover the original input point cloud from the intermediate results. In this paper, we design VVRec, to the best of our knowledge, which is the first targeting DL-based Volumetric Video Reconstruction attack scheme. VVRec demonstrates the ability to reconstruct high-quality point clouds from intercepted transmission intermediate results using four well-trained neural network modules we design. Leveraging the latest latent diffusion models with Gamma distribution and a refinement algorithm, VVRec excels in reconstruction quality, color recovery, and surpasses existing defenses. We evaluate VVRec using three volumetric video datasets. The results demonstrate that VVRec achieves 64.70dB reconstruction accuracy, with an impressive 46.39% reduction of distortion over baselines.

Advancing Language Model Reasoning through Reinforcement Learning and Inference Scaling

Large language models (LLMs) have demonstrated remarkable capabilities in complex reasoning tasks. However, existing approaches mainly rely on imitation learning and struggle to achieve effective test-time scaling. While reinforcement learning (RL) holds promise for enabling self-exploration and learning from feedback, recent attempts yield only modest improvements in complex reasoning. In this paper, we present T1 to scale RL by encouraging exploration and understand inference scaling. We first initialize the LLM using synthesized chain-of-thought data that integrates trial-and-error and self-verification. To scale RL training, we promote increased sampling diversity through oversampling. We further employ an entropy bonus as an auxiliary loss, alongside a dynamic anchor for regularization to facilitate reward optimization. We demonstrate that T1 with open LLMs as its base exhibits inference scaling behavior and achieves superior performance on challenging math reasoning benchmarks. For example, T1 with Qwen2.5-32B as the base model outperforms the recent Qwen QwQ-32B-Preview model on MATH500, AIME2024, and Omni-math-500. More importantly, we present a simple strategy to examine inference scaling, where increased inference budgets directly lead to T1's better performance without any additional verification. We will open-source the T1 models and the data used to train them at \url{https://github.com/THUDM/T1}.

The Essence of Contextual Understanding in Theory of Mind: A Study on Question Answering with Story Characters

Theory-of-Mind (ToM) is a fundamental psychological capability that allows humans to understand and interpret the mental states of others. Humans infer others' thoughts by integrating causal cues and indirect clues from broad contextual information, often derived from past interactions. In other words, human ToM heavily relies on the understanding about the backgrounds and life stories of others. Unfortunately, this aspect is largely overlooked in existing benchmarks for evaluating machines' ToM capabilities, due to their usage of short narratives without global backgrounds. In this paper, we verify the importance of understanding long personal backgrounds in ToM and assess the performance of LLMs in such realistic evaluation scenarios. To achieve this, we introduce a novel benchmark, CharToM-QA, comprising 1,035 ToM questions based on characters from classic novels. Our human study reveals a significant disparity in performance: the same group of educated participants performs dramatically better when they have read the novels compared to when they have not. In parallel, our experiments on state-of-the-art LLMs, including the very recent o1 model, show that LLMs still perform notably worse than humans, despite that they have seen these stories during pre-training. This highlights the limitations of current LLMs in capturing the nuanced contextual information required for ToM reasoning.

How Far is Video Generation from World Model: A Physical Law Perspective

OpenAI's Sora highlights the potential of video generation for developing world models that adhere to fundamental physical laws. However, the ability of video generation models to discover such laws purely from visual data without human priors can be questioned. A world model learning the true law should give predictions robust to nuances and correctly extrapolate on unseen scenarios. In this work, we evaluate across three key scenarios: in-distribution, out-of-distribution, and combinatorial generalization. We developed a 2D simulation testbed for object movement and collisions to generate videos deterministically governed by one or more classical mechanics laws. This provides an unlimited supply of data for large-scale experimentation and enables quantitative evaluation of whether the generated videos adhere to physical laws. We trained diffusion-based video generation models to predict object movements based on initial frames. Our scaling experiments show perfect generalization within the distribution, measurable scaling behavior for combinatorial generalization, but failure in out-of-distribution scenarios. Further experiments reveal two key insights about the generalization mechanisms of these models: (1) the models fail to abstract general physical rules and instead exhibit "case-based" generalization behavior, i.e., mimicking the closest training example; (2) when generalizing to new cases, models are observed to prioritize different factors when referencing training data: color > size > velocity > shape. Our study suggests that scaling alone is insufficient for video generation models to uncover fundamental physical laws, despite its role in Sora's broader success. See our project page at https://phyworld.github.io

AdaNAT: Exploring Adaptive Policy for Token-Based Image Generation

Recent studies have demonstrated the effectiveness of token-based methods for visual content generation. As a representative work, non-autoregressive Transformers (NATs) are able to synthesize images with decent quality in a small number of steps. However, NATs usually necessitate configuring a complicated generation policy comprising multiple manually-designed scheduling rules. These heuristic-driven rules are prone to sub-optimality and come with the requirements of expert knowledge and labor-intensive efforts. Moreover, their one-size-fits-all nature cannot flexibly adapt to the diverse characteristics of each individual sample. To address these issues, we propose AdaNAT, a learnable approach that automatically configures a suitable policy tailored for every sample to be generated. In specific, we formulate the determination of generation policies as a Markov decision process. Under this framework, a lightweight policy network for generation can be learned via reinforcement learning. Importantly, we demonstrate that simple reward designs such as FID or pre-trained reward models, may not reliably guarantee the desired quality or diversity of generated samples. Therefore, we propose an adversarial reward design to guide the training of policy networks effectively. Comprehensive experiments on four benchmark datasets, i.e., ImageNet-256 & 512, MS-COCO, and CC3M, validate the effectiveness of AdaNAT. Code and pre-trained models will be released at https://github.com/LeapLabTHU/AdaNAT.