<aside> 🔬
BandAI is dedicated to exploring next-generation products and technologies for Douyin E-commerce, Local Services, and Advertising, as well as the most challenging long-term key problems in the field of LLMs, leveraging technological breakthroughs to deliver unprecedented user experiences.
Research directions include:
Join us — help ensure AI creates real-world utility.
</aside>
Table of Contents
In real-world workflows, critical information is often stored not as clean text but across heterogeneous files such as spreadsheets, reports, documents, and slide decks, where evidence may be hidden in multi-sheet layouts, merged cells, formulas, hierarchical structures, or embedded objects. To address this challenge, we introduce File Reasoning, a setting where agents must interact directly with raw XLSX, PDF, DOCX, and PPTX files in a persistent sandbox, using executable tools such as Python and Shell commands to inspect file structure, extract evidence, recover from errors, and produce verifiable answers. Based on this setting, we construct a challenging benchmark of 410 verified file-grounded questions and propose FIRE, a two-stage training framework that first teaches models basic tool-use behavior through supervised fine-tuning and then improves long-horizon planning through stateful reinforcement learning with real execution feedback. Experiments show that FIRE substantially improves open-source models on real-world file reasoning tasks, with Qwen3-32B-FIRE achieving the best performance among comparable open-source models, demonstrating that environment-grounded, stateful training is an effective path toward more capable file-based agents.
Poster generation is a complex task demanding a harmonious integration of visual aesthetics and information hierarchy. While recent textto-image models have advanced visual synthesis, they remain non-editable and struggle with precise text rendering. Conversely, existing layoutgeneration methods offer structure but typically rely on static, one-shot predictions, lacking a mechanism for self-correction essential to professional design. Inspired by the iterative workflow of human designers, we introduce PosterAgent, a novel framework that reformulates poster creation as an agentic workflow involving initial drafting followed by iterative refinement. To effectively train this multi-turn capability, we propose Stage-Aware Reinforcement Learning (SARL), which decouples the optimization into draft-specific and refinement-specific phases, ensuring precise credit assignment for both initial drafting and incremental refinement actions. Extensive experiments demonstrate that PosterAgent significantly outperforms strong baselines, validating the potential of agentic systems in graphic design.
Achieving mastery in real world software engineering tasks is fundamentally bottlenecked by the scarcity of large scale, high quality training data. Scaling such data has been limited by the complexity of environment setup, unit test generation, and problem statement curation. In this paper, we propose ScaleSWE, an automated, sandboxed multi agent workflow designed to construct high quality SWE data at scale. The system coordinates three specialized agents for environment setup, test creation, and problem description synthesis to process 6 million pull requests across 5200 repositories, producing Scale SWE Data: 100k verified SWE instances, the largest such dataset to date. It substantially surpasses existing real world datasets in repository diversity and reflects realistic task complexity. We further demonstrate the dataset utility for training by distilling 71498 high quality trajectories and finetuning Qwen30BA3BInstruct to produce ScaleSWE Agent. Our agent achieves a 64 resolve rate on SWE Bench Verified a nearly three fold improvement over the base model. ScaleSWE provides a scalable, reproducible approach for data construction to advance LLM based software engineering. Scale SWE will be publicly available.
Reinforcement Learning (RL) has empowered Large Language Models (LLMs) with strong reasoning capabilities, but vanilla RL mainly focuses on generation capability improvement by training with only first-order rollout (generating multiple responses for a question), and we argue that this approach fails to fully exploit the potential of training data because of the neglect of critique capability training. To tackle this problem, we further introduce the concept of second-order rollout (generating multiple critiques for a response) and propose a unified framework for jointly training generation and critique capabilities. Extensive experiments across various models and datasets demonstrate that our approach can utilize training data more effectively than vanilla RL and achieve better performance under the same training data. Additionally, we uncover several insightful findings regarding second-order rollout and critique training, such as the importance of label balance in critique training and the noise problem of outcome-based rewards, which can be mitigated through sampling techniques. Our work offers a preliminary exploration of dynamic data augmentation and joint generation-critique training in RL, providing meaningful inspiration for the further advancement of RL training.
Large language models (LLMs) show promise as teaching assistants, yet their teaching capability remains insufficiently evaluated. Existing benchmarks mainly focus on problem-solving or problem-level guidance, leaving knowledge-centered teaching underexplored. We propose a syllabus-grounded evaluation framework that measures LLM teaching capability via student performance improvement after multi-turn instruction. By restricting teacher agents to structured knowledge points and example problems, the framework avoids information leakage and enables reuse of existing benchmarks.
We instantiate the framework on Gaokao data across multiple subjects. Experiments reveal substantial variation in teaching effectiveness across models and domains: some models perform well in mathematics, while teaching remains challenging in physics and chemistry. We also find that incorporating example problems does not necessarily improve teaching, as models often shift toward example-specific error correction. Overall, our results highlight teaching ability as a distinct and measurable dimension of LLM behavior.
Chain-of-Thought (CoT) is a critical technique in enhancing the reasoning ability of Large Language Models (LLMs), and latent reasoning methods have been proposed to accelerate the inefficient token-level reasoning chain. We notice that existing latent reasoning methods generally require model structure augmentation and exhaustive training, limiting their broader applicability. In this paper, we propose CoLT, a novel framework that implements latent reasoning as “tool calls”. Instead of reasoning entirely in the latent space, CoLT generates seed tokens that contain information of a reasoning step.When a latent tool call is triggered, a smaller external model will take the hidden states of seed tokens as its input, and unpack the seed tokens back to a full reasoning step. In this way, we can ensure that the main model reasons in the explicit token space, preserving its ability while improving efficiency. Experimental results on four mathematical datasets demonstrate that CoLT achieves higher accuracy and shorter reasoning length than baseline latent models, and is compatible with reinforcement learning algorithms and different decoder structures.