Dalian University of Technology

7 related articles

Subscribe to updates

arXiv Oct 24, 2026
Patent review gets a trainable agent testbed: a 4B model holds its own against far larger general agents at prior-art search
Patent examination asks an agent to find sparse prior art, refine queries, inspect evidence and write a novelty judgment, yet existing benchmarks mostly stop at static candidate sets. PatEx-Gym turns that process into a runnable environment: PatEx-Bench supplies 1K examination tasks, retrieval corpora of over 1M patents and 52K scientific papers, and a runtime supporting iterative search and Office Action generation. PASS synthesizes search trajectories by treating document visits as implicit relevance signals, and the 4B PatEx-Agent fine-tuned on them performs competitively against substantially larger general-purpose agents.
Patent examination asks an agent to find sparse prior art, refine queries, inspect evidence and write a novelty judgment, yet existing benchmarks mostly stop at static candidate sets. PatEx-Gym turns that process into a runnable environment: PatEx-Bench supplies 1K examination tasks, retrieval corpora of over 1M patents and 52K scientific papers, and a runtime supporting iterative search and Office Action generation. PASS synthesizes search trajectories by treating document visits as implicit relevance signals, and the 4B PatEx-Agent fine-tuned on them performs competitively against substantially larger general-purpose agents.
Dalian University of Technology Chinese Academy of Sciences Shenzhen University of Advanced Technology
Qiyao Wang（First author）
PatEx-Gym Agentic Search Patent Examination
researchsquare.com Oct 19, 2026
Grid conical shell milling gets a physics chain: predict cutting force first, then compute deflection, with accuracy above 93% on all four test sets
When a thin-walled grid conical shell is milled, the rib layout makes local stiffness jump from tool position to tool position, and the cutting force pushes the workpiece into an elastic retreat along the shell normal, so the actual wall thickness falls short of the theoretical depth of cut. The new route treats finite-element simulation as the source domain and real milling experiments as the target domain, uses transfer learning to predict the normal cutting force first, then feeds that load into an equivalent-stiffness analytical model to compute deflection. Across four test sets the method reached 93.05% to 95.27% prediction accuracy, above the CNN, PINN and TL comparison models.
When a thin-walled grid conical shell is milled, the rib layout makes local stiffness jump from tool position to tool position, and the cutting force pushes the workpiece into an elastic retreat along the shell normal, so the actual wall thickness falls short of the theoretical depth of cut. The new route treats finite-element simulation as the source domain and real milling experiments as the target domain, uses transfer learning to predict the normal cutting force first, then feeds that load into an equivalent-stiffness analytical model to compute deflection. Across four test sets the method reached 93.05% to 95.27% prediction accuracy, above the CNN, PINN and TL comparison models.
Dalian University of Technology
Xingliang Chai（First author）
PCTD Grid conical shell Transfer Learning Equivalent stiffness model Elastic deformation prediction
arXiv Oct 8, 2026
SLR supervises continuous latent reasoning with a fingertip ray state plus parity-pooled visual targets, raising EgoPoint-Ground mIoU over same-backbone SFT by up to 21.1 points
The work proposes Spatial Latent Reasoning (SLR), which organizes pointing-gesture geometry and target-region appearance into an ordered sequence of continuous latent states: one spatial ray state supervised by fingertip position and pointing direction, followed by four states aligned with target-region features whose visual targets are built by parity pooling, which groups ROI tokens by row and column parity and averages each group; on EgoPoint-Ground it improves mIoU over same-backbone supervised fine-tuning by 2.8, 17.5, and 21.1 percentage points on Qwen3.5-4B, Qwen2.5-VL-7B, and Qwen3-VL-8B, with gains on both hard subsets, and reaches 77.6% precision at IoU 0.5 on YouRefIt.
The work proposes Spatial Latent Reasoning (SLR), which organizes pointing-gesture geometry and target-region appearance into an ordered sequence of continuous latent states: one spatial ray state supervised by fingertip position and pointing direction, followed by four states aligned with target-region features whose visual targets are built by parity pooling, which groups ROI tokens by row and column parity and averages each group; on EgoPoint-Ground it improves mIoU over same-backbone supervised fine-tuning by 2.8, 17.5, and 21.1 percentage points on Qwen3.5-4B, Qwen2.5-VL-7B, and Qwen3-VL-8B, with gains on both hard subsets, and reaches 77.6% precision at IoU 0.5 on YouRefIt.
Dalian University of Technology Tsinghua University
Ling Li（First author）
Spatial Latent Reasoning Parity Pooling Embodied Reference Understanding Visual Grounding Latent Visual Reasoning
arXiv Oct 5, 2026
MITE instruction-tunes with Python, C++, and Java code formats and ensembles by voting, consistently beating BERT and LLM baselines on six BioNER datasets
The work proposes MITE, which reformulates biomedical named entity recognition as a structure-to-structure generation task by rendering each training instance in multiple programming-language formats such as Python, C++, and Java to supply structurally diverse supervision without external biomedical knowledge or extra annotations, and at inference aggregates predictions from the different code formats through entity-level voting; on six widely used BioNER datasets MITE consistently outperforms representative BERT-based and LLM-based baselines and shows strong cross-dataset generalization, with ablation and parameter analyses further supporting the effectiveness and robustness of the proposed components.
The work proposes MITE, which reformulates biomedical named entity recognition as a structure-to-structure generation task by rendering each training instance in multiple programming-language formats such as Python, C++, and Java to supply structurally diverse supervision without external biomedical knowledge or extra annotations, and at inference aggregates predictions from the different code formats through entity-level voting; on six widely used BioNER datasets MITE consistently outperforms representative BERT-based and LLM-based baselines and shows strong cross-dataset generalization, with ablation and parameter analyses further supporting the effectiveness and robustness of the proposed components.
Dalian Maritime University Dalian University of Technology Northeastern University
Songtao Li（First author）
MITE Biomedical Named Entity Recognition Instruction Tuning
arXiv Oct 2, 2026
Kuaishou team's DARA weights rewards by active-group density, cutting tool-calling format steps by up to 26% and math length-compliance steps by up to 65%
The work characterizes each reward's contribution to policy updates in multi-reward RL through "advantage energy," shows that under idealized GDPO normalization a reward's advantage energy is proportional to its active-group density (the fraction of rollout groups in which it yields nonzero relative advantages), and proposes DARA, an inverse-square-root density weighting that reaches target behaviors faster than GDPO: up to 26% fewer training steps for format compliance in tool calling and up to 65% fewer steps for near-saturated length compliance in mathematical reasoning, while remaining competitive in final performance.
The work characterizes each reward's contribution to policy updates in multi-reward RL through "advantage energy," shows that under idealized GDPO normalization a reward's advantage energy is proportional to its active-group density (the fraction of rollout groups in which it yields nonzero relative advantages), and proposes DARA, an inverse-square-root density weighting that reaches target behaviors faster than GDPO: up to 26% fewer training steps for format compliance in tool calling and up to 65% fewer steps for near-saturated length compliance in mathematical reasoning, while remaining competitive in final performance.
Beijing Foreign Studies University University of Minnesota University of Wisconsin–Madison Dalian University of Technology Kuaishou (China)Southeast University Stony Brook University
Tong Zheng（First author）
Density-Aware Reward Aggregation Advantage energy Multi-reward reinforcement learning tool calling Mathematical Reasoning
arXiv Sep 30, 2026
APM-Bench tests cross-session persistent memory for streaming video assistants across 549 sessions and 104 life trajectories, finding existing methods struggle to combine recall, latency, and storage
The work introduces APM-Bench, which reformulates egocentric streaming interaction as multi-session life trajectories (549 sessions, 104 trajectories, 2,719 candidates, averaging 69 minutes of video per trajectory) and evaluates general video models and eight specialized memory systems on cross-session understanding, real-time perception, and adaptive response, plus an evidence-availability-aware test; results reveal a clear utility-latency-storage trade-off: raw video memory gives the strongest cross-session performance but needs GiB-scale storage and high latency, text summaries cut storage to the KiB scale at the cost of cross-session performance, event-structured memory shows the strongest utility among specialized systems, and adaptive response remains difficult even with rich history
The work introduces APM-Bench, which reformulates egocentric streaming interaction as multi-session life trajectories (549 sessions, 104 trajectories, 2,719 candidates, averaging 69 minutes of video per trajectory) and evaluates general video models and eight specialized memory systems on cross-session understanding, real-time perception, and adaptive response, plus an evidence-availability-aware test; results reveal a clear utility-latency-storage trade-off: raw video memory gives the strongest cross-session performance but needs GiB-scale storage and high latency, text summaries cut storage to the KiB scale at the cost of cross-session performance, event-structured memory shows the strongest utility among specialized systems, and adaptive response remains difficult even with rich history
Hong Kong Polytechnic University Beihang University Shanghai Jiao Tong University Dalian University of Technology Chinese Academy of Sciences Eastern Institute of Technology, Ningbo
Jianguo Huang（First author）
APM-Bench Persistent Memory Streaming Video Understanding
arXiv Sep 21, 2026
Foresight Without Trajectories: Guiding 3D Diffusion Policies with a Latent Movement Trend
The work introduces Movement Trend Guidance: a compact latent of interaction evolution is learned from a short history of point clouds and robot states, supervised during training only by sparse future gripper states, and at inference retained alone as future-oriented conditioning, consistently improving 3D diffusion policies on RoboTwin2.0, LIBERO-40, DexArt and five real-robot tasks while adding only 3.52% more parameters than DP3.
The work introduces Movement Trend Guidance: a compact latent of interaction evolution is learned from a short history of point clouds and robot states, supervised during training only by sparse future gripper states, and at inference retained alone as future-oriented conditioning, consistently improving 3D diffusion policies on RoboTwin2.0, LIBERO-40, DexArt and five real-robot tasks while adding only 3.52% more parameters than DP3.
Dalian University of Technology
Zhongbo Zhang（First author）
Movement Trend Guidance 3D Diffusion Policy RoboTwin2.0 Robotic Manipulation

You’re all caught up

Patent review gets a trainable agent testbed: a 4B model holds its own against far larger general agents at prior-art search

Grid conical shell milling gets a physics chain: predict cutting force first, then compute deflection, with accuracy above 93% on all four test sets

SLR supervises continuous latent reasoning with a fingertip ray state plus parity-pooled visual targets, raising EgoPoint-Ground mIoU over same-backbone SFT by up to 21.1 points

MITE instruction-tunes with Python, C++, and Java code formats and ensembles by voting, consistently beating BERT and LLM baselines on six BioNER datasets

Kuaishou team's DARA weights rewards by active-group density, cutting tool-calling format steps by up to 26% and math length-compliance steps by up to 65%

APM-Bench tests cross-session persistent memory for streaming video assistants across 549 sessions and 104 life trajectories, finding existing methods struggle to combine recall, latency, and storage

Foresight Without Trajectories: Guiding 3D Diffusion Policies with a Latent Movement Trend