Cheng Ren

ByteDance Seed, Jiaze Chen, Tiantian Fan et al. · bytedance

We introduce Seed1.5-Thinking, capable of reasoning through thinking before responding, resulting in improved performance on a wide range of benchmarks. Seed1.5-Thinking achieves 86.7 on AIME 2024, 55.0 on Codeforces and 77.3 on GPQA, demonstrating excellent reasoning abilities in STEM and coding. Beyond reasoning tasks, the method demonstrates notable generalization across diverse domains. For instance, it surpasses DeepSeek R1 by 8% in win rate on non-reasoning tasks, indicating its broader applicability. Compared to other state-of-the-art reasoning models, Seed1.5-Thinking is a Mixture-of-Experts (MoE) model with a relatively small size, featuring 20B activated and 200B total parameters. As part of our effort to assess generalized reasoning, we develop two internal benchmarks, BeyondAIME and Codeforces, both of which will be publicly released to support future research. Model trial link: https://www.volcengine.com/experience/ark.

Cheng Ren, Ming Li, Xinping Guan et al.

Robotic Cellular Warehousing Systems (RCWS) give rise to multi-agent pickup and delivery (MAPD) processes in which robots sequentially collect multiple stock-keeping units (SKUs) for each order. Unlike classical MAPD formulations that assume static tasks, real warehouse operations often involve dynamic order evolution, where new SKUs may be appended to an order while it is being executed. Motivated by this practical requirement, this letter formulates the Dynamic Multi-Agent Pickup and Delivery problem considering internal order evolution for the first time. Building on the token passing paradigm, we propose two event-triggered online replanning algorithms. The first, Dynamic Token Passing, performs localized replanning upon order updates through add-order decomposition and priority-based token scheduling while preserving collision-free execution. The second, Cooperative Token Passing, further enables idle robots to opportunistically assist newly added pickups, improving system-level efficiency. Simulation results in RCWS environments demonstrate that the proposed methods significantly reduce order flowtime compared with static and non-cooperative baselines.

Li Lin, Yamini Sri Krubha, Zhenhuan Yang et al.

In the realm of medical imaging, particularly for COVID-19 detection, deep learning models face substantial challenges such as the necessity for extensive computational resources, the paucity of well-annotated datasets, and a significant amount of unlabeled data. In this work, we introduce the first lightweight detector designed to overcome these obstacles, leveraging a frozen CLIP image encoder and a trainable multilayer perception (MLP). Enhanced with Conditional Value at Risk (CVaR) for robustness and a loss landscape flattening strategy for improved generalization, our model is tailored for high efficacy in COVID-19 detection. Furthermore, we integrate a teacher-student framework to capitalize on the vast amounts of unlabeled data, enabling our model to achieve superior performance despite the inherent data limitations. Experimental results on the COV19-CT-DB dataset demonstrate the effectiveness of our approach, surpassing baseline by up to 10.6% in `macro' F1 score in supervised learning. The code is available at https://github.com/Purdue-M2/COVID-19_Detection_M2_PURDUE.

Luoxin Chen, Jinming Gu, Liankai Huang et al. · cmu

LLMs have demonstrated strong mathematical reasoning abilities by leveraging reinforcement learning with long chain-of-thought, yet they continue to struggle with theorem proving due to the lack of clear supervision signals when solely using natural language. Dedicated domain-specific languages like Lean provide clear supervision via formal verification of proofs, enabling effective training through reinforcement learning. In this work, we propose \textbf{Seed-Prover}, a lemma-style whole-proof reasoning model. Seed-Prover can iteratively refine its proof based on Lean feedback, proved lemmas, and self-summarization. To solve IMO-level contest problems, we design three test-time inference strategies that enable both deep and broad reasoning. Seed-Prover proves $78.1\%$ of formalized past IMO problems, saturates MiniF2F, and achieves over 50\% on PutnamBench, outperforming the previous state-of-the-art by a large margin. To address the lack of geometry support in Lean, we introduce a geometry reasoning engine \textbf{Seed-Geometry}, which outperforms previous formal geometry engines. We use these two systems to participate in IMO 2025 and fully prove 5 out of 6 problems. This work represents a significant advancement in automated mathematical reasoning, demonstrating the effectiveness of formal verification with long chain-of-thought reasoning.

Borui Wan, Gaohong Liu, Zuquan Song et al.

The training scale of large language models (LLMs) has reached tens of thousands of GPUs and is still continuously expanding, enabling faster learning of larger models. Accompanying the expansion of the resource scale is the prevalence of failures (CUDA error, NaN values, job hang, etc.), which poses significant challenges to training stability. Any large-scale LLM training infrastructure should strive for minimal training interruption, efficient fault diagnosis, and effective failure tolerance to enable highly efficient continuous training. This paper presents ByteRobust, a large-scale GPU infrastructure management system tailored for robust and stable training of LLMs. It exploits the uniqueness of LLM training process and gives top priorities to detecting and recovering failures in a routine manner. Leveraging parallelisms and characteristics of LLM training, ByteRobust enables high-capacity fault tolerance, prompt fault demarcation, and localization with an effective data-driven approach, comprehensively ensuring continuous and efficient training of LLM tasks. ByteRobust is deployed on a production GPU platform and achieves 97% ETTR for a three-month training job on 9,600 GPUs.

Chen Zheng, Yiyuan Ma, Yuan Yang et al.

The development of alignment and reasoning capabilities in large language models has seen remarkable progress through two paradigms: instruction tuning and reinforcement learning from human feedback (RLHF) alignment paradigm, and distillation-based reasoning fine-tuning paradigm. While both approaches prove effective independently, the third paradigm of applying RLHF to distillation-trained models presents significant challenges. Our investigation reveals two critical phenomena that emerge in this paradigm: Sequence Length Collapse, where language generation dramatically reduces during early RLHF training, and the Reward Hockey Stick Curve, featuring severe reward score drops followed by gradual recovery. These instabilities fundamentally compromise the model's alignment and reasoning capabilities. To address these challenges, we propose Balanced Actor Initialization (BAI), a two-stage weighted model merging approach. BAI first merges instruction-following and distillation-based reasoning fine-tuned models, then further combines this intermediate model with the pretrained model to preserve foundational knowledge. Through comprehensive experiments across diverse benchmarks and detailed analysis of training experiments, we demonstrate that BAI resolves Sequence Length Collapse, mitigates the Reward Hockey Stick Curve, and enables continuous sequence length improvement during training. Additionally, our analysis reveals that balanced merging ratios achieve optimal trade-offs between training stability and reasoning capability preservation. Our work provides the effective solution for stable training in this third paradigm, enabling more capable reasoning models that combine distillation efficiency with RLHF alignment.

Zhi Li, Cheng Ren, Xianyou Li et al.

Assessment and reporting of skills is a central feature of many digital learning platforms. With students often using multiple platforms, cross-platform assessment has emerged as a new challenge. While technologies such as Learning Tools Interoperability (LTI) have enabled communication between platforms, reconciling the different skill taxonomies they employ has not been solved at scale. In this paper, we introduce and evaluate a methodology for finding and linking equivalent skills between platforms by utilizing problem content as well as the platform's clickstream data. We propose six models to represent skills as continuous real-valued vectors and leverage machine translation to map between skill spaces. The methods are tested on three digital learning platforms: ASSISTments, Khan Academy, and Cognitive Tutor. Our results demonstrate reasonable accuracy in skill equivalency prediction from a fine-grained taxonomy to a coarse-grained one, achieving an average recall@5 of 0.8 between the three platforms. Our skill translation approach has implications for aiding in the tedious, manual process of taxonomy to taxonomy mapping work, also called crosswalks, within the tutoring as well as standardized testing worlds.