Yiqin Hao, Boyi Zeng, Zitong Wang, He Li, Shixiang Song, Feichen Song, Yifan Liu, Ran Tao 等 10 位作者
Mathematical problem solving and autonomous agents are increasingly important applications of language models, demanding stronger reasoning capabilities. Computational depth is a key resource for such reasoning, but adding model layers increases both training and inference costs, while test-time scaling incurs substantial additional inference computation. We introduce P2N (Previous to Next), which reuses the previous token's deep representations, already available during generation, to increase effective computational depth with negligible decoding overhead. Specifically, P2N adds the previous token's deep hidden state to the current token's shallow representation, allowing subsequent layers to build on previously completed deep computation. This recurrence extends effective computational depth across tokens, while introducing only a single vector addition per decoding step. We train the recurrence using parallel Jacobi updates restricted to a middle-layer core. Experiments with Qwen-style models show that, at 1.5B parameters, P2N improves GSM8K accuracy by over 10 percentage points under both zero-shot and few-shot prompting, with nearly unchanged decoding latency. The benefits also extend to language modeling and average performance on general downstream benchmarks. on general downstream benchmarks.
Ziyang Zeng, Tingyang Chen, Jia Deng, Qiyao Wang, Shuo Tang, Yike Yang, Tao Ren, Wayne Xin Zhao 等 13 位作者
Scaling supervised fine-tuning for working agents requires not only broad coverage of professional tasks, but also expert demonstrations that faithfully carry out professional workflows to completion. However, missing or inconsistent task information can disrupt expert demonstrations, weakening the resulting supervision. We introduce WorkFactory, a scalable task-construction framework that turns professional intents into executable task environments through hybrid intent sourcing, semantic workflow planning, and execution-based repair. Starting from these intents, WorkFactory constructs a Semantic Workflow Graph that explicitly represents information requirements, dependencies, and locations within the task environment. The graph guides the instantiation of task environments, while an adversarial execution-refinement loop exposes missing or inconsistent information and steers targeted repairs that preserve the original intent. Using WorkFactory, we construct 112,686 tasks spanning 612 occupations and train the WorkPilot family at 4B, 27B, and 35B-A3B scales on the resulting teacher trajectories. All three models improve over their respective base models across four professional-work benchmarks, with WorkPilot-35B-A3B rising from 879 to 1565 Elo on GDPval-AA v2.1. Scaling experiments with up to 110k tasks further show that larger models benefit more from additional task data but exhibit diminishing returns at smaller data scales, while WorkPilot's long-horizon execution evolves from verbose trajectories toward more efficient task completion. Controlled component studies quantify the negative effects of task-environment defects on downstream performance and show how semantic workflow planning and execution-based repair yield downstream gains through improved task-environment quality. The subset of data are available at https://github.com/IQuestLab/WorkFactory.
Long-horizon LLM agents interact with their environments through complete assistant turns while generating each action token by token. We observe that critic values vary primarily across turns, motivating optimization anchored at turns with finer credit allocation among tokens. We propose Multiscale Policy Optimization (MPO), combining Multiscale Advantage Estimation (MAE) and turn-level importance sampling to align credit assignment and policy updates at the turn scale. Extensive experiments across search, tool-use, and automation tasks with 4B and 30B-A3B models demonstrate consistent gains over strong RL baselines.
We present Reap, an open framework for learning and analyzing formal proof search in Lean4. Reap implements policy-value-guided Monte Carlo Tree Search directly in Lean, sharing proof-state transitions, AND-OR search, and replay verification between inference and training rollouts. Structured action logs expose how neural predictions and tactic execution interact during search. Using this framework, we train Reaper-1, a 1.7B-parameter policy-value model, through continued pre-training, supervised tactic learning, and iterative search-based reinforcement learning. Using a separate premise selector, the final system solves 79.9% of miniF2F-test at pass@32, establishing state-of-the-art performance among open provers of comparable size while remaining competitive with substantially larger models. Beyond improved solve rates, our results suggest that RL strengthens the prover's ability to organize proofs around useful intermediate results, favoring substantial advances over repeated local adjustments. This more directed proof construction is accompanied by increasingly elaborate tactic programs: fewer explicit search decisions are needed, while more work is performed within each decision. The resulting trade-off between search effort and tactic execution cost provides insight into how learning changes the prover's problem-solving strategy.
Zitian Gao, Yilong Chen, Yihao Xiao, Xinyu Yang, Ran Tao, Joey Zhou, Bryan Dai
We present Loopie, the most powerful open-source looped Transformer to date. The Loopie series consists of two Mixture-of-Experts (MoE) models: a 20B-parameter model with 2B active parameters and a 6B parameter model with 0.6B active parameters. Looped Transformers have long faced a challenge: given an N-fold pre-training compute, increasing the parameter count by a factor of N usually outperforms looping N times. Loopie addresses this challenge. Extensive ablation studies, including comparisons with a vanilla 30B-A3B model, show that Loopie substantially outperforms vanilla Transformer baselines trained with the same compute budget. With a novel post-training method, Loopie develops strong reasoning abilities and achieves frontier-level reasoning performance.
Yunqi Liu, Peiyu Wang, Bingli Wang, Leo Luo, Xuanwu Song, Xudong Ma, Hongxin Li, Yaxin Du 等 24 位作者
Command-line interface (CLI) agents increasingly execute longer-horizon tasks, yet existing benchmarks either assume that user instructions remain fixed throughout execution or evaluate task outcomes without capturing runtime collaboration quality. In this paper, we present REALCLI, a paired benchmark that unifies the evaluation of task execution and human-agent collaboration quality on realistic CLI tasks. To comprehensively measure collaboration, REALCLI builds a five-dimensional framework, operationalized into 22 sub-dimensions and 51 interaction scenario templates. REALCLI provides two modes for paired comparison: REALCLI-EXEC supplies complete specifications upfront, whereas REALCLI-COLLAB introduces a scripted user that delivers progressive information, corrections, constraint revisions, and authorization decisions. Evaluations reveal profound capability gaps in frontier agents: the best configuration attains an average score of only 49.2%, and when evaluated by the joint collaboration score ($J_{\mathrm{collab}}$), the peak performance reaches merely 37.6%. These results show that task effectiveness diverges sharply from collaboration quality, with models consistently struggling in steerability, trust calibration, and safety boundary management. REALCLI provides a rigorous, reproducible foundation for measuring and advancing collaborative CLI agents.
Yilong Chen, Zitian Gao Ubiquant, Yihao Xiao, Jason Klein Liu, Xinyu Yang Carnegie Mellon University, Yifan Luo Peking University, Haoming Luo, Zhengmao Ye 等 11 位作者
Although the Universal Transformer (UT) mitigates the diminishing returns of standard LLM scaling by decoupling parameter count from depth, it remains constrained by linear computational costs and rigid weight-sharing mechanisms. These limitations lead to severe functional homogeneity, which subsequently induces over-smoothing, representation rank collapse, and degraded reasoning performance. In this work, we present the first systematic study of Compute Distribution Skew, identifying it as the primary driver of extrapolation failure. This is a pathological phenomenon in ultra-deep recurrent Transformers characterized by a disproportionate distribution of contributions across recurrent steps, resulting in distinct functional states during prefix and suffix processing phases. To address this challenge, we propose the Polymorphic Transformer, which aims to achieve functional polymorphism and depth sparsity within a shared-parameter framework. By integrating conditional sparse subspaces, SiLU Attention, and an uncertainty-aware depth scheduler, our architecture mitigates power-method collapse and effectively decouples logical depth from computational cost. Experiments demonstrate that our model significantly enhances representation rank and robustness, achieving complex reasoning performance comparable to baseline while reducing computation by 64.7%.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表