Skip to main content
Search

Chinese Academy of Sciences

71 related articles

Subscribe to updates
  1. Advanced Science

    Enzyme function prediction learns from unlabeled homologs: EZpred lifts four-digit EC F1 to 0.786 on 720 newly annotated enzymes

    EZpred carries the homolog-based features long used in protein structure prediction into enzyme function prediction, taking only sequence features from homologs and ignoring their function labels. Most deep learning models for this task read features from the target sequence alone, and the few that use homologs only aggregate labeled ones. On 720 enzymes with new EC numbers in UniProt 2024_02, EZpred reaches an F1 of 0.786 for full four-digit EC numbers, 27% and 31% above CLEAN and GraphEC.

  2. arXiv

    Agent RL stops scoring whole trajectories: ORSD writes its own rubrics to assign credit token by token

    Online Rubric Self-Distillation (ORSD) has the current policy generate natural-language process rubrics from differences among a group of its own trajectories, lets a frozen verifier pick the bundle whose rubric satisfaction best correlates with task success, and uses that bundle as privileged context: the log-likelihood gap for the same tokens with and without the rubric becomes a token-level correction to the trajectory-level GRPO advantage. On ALFWorld and WebShop, Qwen3-1.7B and Qwen3-4B reach the best overall results among the compared methods under the same environment-interaction budget.

  3. arXiv

    CLEAR turns papers into checkable execution specs first: 29.87 on PaperBench at $7.91 per paper

    CLEAR inserts a structured execution-specification layer between a research paper and runnable code, converting research documents into human-checkable implementation units, dependencies, interfaces and evaluation procedures before generating and locally repairing code. On 20 PaperBench tasks it averages a score of 29.87 at an average cost of $7.91 per paper; against AiScientist, cost falls 35.16% while the score gives up 3.86 points. On MLE-Bench Lite, all 22 tasks produce valid submissions and 77.27% win any medal.

  4. arXiv

    Multi-turn RL pushes LLM agents past the base model only with roomy budgets, and the edge fades or reverses when turns tighten

    Multi-turn agentic reinforcement learning is widely used to strengthen the long-horizon behavior of large language model agents, yet trajectory-level success rates hide what it does to individual decisions. Across four interactive tasks, DeepCoder-MT, BFCL, InterCode-SQL and FrozenLake, models trained with GRPO are evaluated along two axes, sampling breadth and interaction depth: with more sampled trajectories the RL model consistently beats its base model, but when the turns allowed per trajectory are tightened that edge shrinks, disappears, and on some tasks reverses. Matched-state comparisons show better next-step actions on only 46.8% of DeepCoder-MT states and 42.8% of FrozenLake states, with the gain concentrated in error recovery and later trajectory stages.

  5. PLoS ONE

    Thirty heterogeneous control tasks share one sequence backbone: HeteroGrad-DT lifts average normalized return by 32.3%

    In heterogeneous multi-task offline reinforcement learning, tasks differ in state dimension, action space and dynamics, so a directly shared backbone hits both interface mismatch and gradient conflict. HeteroGrad-DT gives each task its own input tokenizer and prediction head around a shared sequence backbone, and a Gradient Flow Transformer pools per-task gradients over a synchronization interval and learns fusion weights to coordinate the shared update. On the DeepMind Control Suite with up to 30 tasks, its average normalized return exceeds representative multi-task methods, with a 32.3% relative gain over standard shared-backbone training and 8.6% over a PCGrad baseline using the same task-aware interfaces.

  6. arXiv

    Retinal cross-modal registration skips segmentation: MICRO benchmark and one-step OFS2Reg reach 0.784 success rate

    Cross-modal retinal image rigid registration has largely relied on a two-step pipeline that segments vessels first and detects keypoints second, so registration quality is capped by segmentation accuracy. The MICRO benchmark gathers 1,365 cross-modal retinal image pairs spanning preoperative, intraoperative and postoperative surgery phases, each with expert-annotated control points. Built on that benchmark, the one-step OFS2Reg constructs keypoint topology constraints directly through a frequency-guided structure preservation descriptor and cross-modal vessel structure similarity learning, reaching a 0.784 success rate and 5.08-pixel RMSE on the MICRO-pre subset.

  7. Solar filament fragments stop being stitched by distance alone: a graph network treats each fragment as a node and reaches AUC 0.867 at kNN 10

    Solar filaments in Hα images break into fragments as they evolve, and projection effects can separate pieces of one filament while bringing pieces of different filaments close together, so distance and orientation alone cannot decide which fragments belong together. The study treats each fragment as a graph node and each candidate pair as an edge, predicts at the edge level with a graph neural network, and replaces conventional additive aggregation with multiplicative aggregation so that inconsistent neighbors contribute less. On a dataset built from about 100 Kanzelhöhe Observatory Hα images and about 1,700 fragments, the method reaches AUC 0.867 and AP 0.792 at kNN 10, above GCN at 0.832/0.747 and GAT at 0.842/0.766.

  8. Graph neural network maps PBE formation energies onto the r2SCAN scale, cutting the cross-functional gap to 14.3 meV/atom

    Formation energies from different exchange-correlation functionals have not been directly mergeable: in MatPES the same structures differ by 107 meV/atom on average between PBE and r2SCAN. Using more than 380,000 structurally paired PBE–r2SCAN entries, researchers trained a graph neural network to predict the energy residual between the two functionals and correct PBE formation energies onto the r2SCAN scale. With GPTFF as the backbone, the model reaches a mean absolute error of 14.3 meV/atom on the held-out test set, below the 18.2 meV/atom of CHGNet.

  9. arXiv

    Disaster reports stop at masks no longer: DisasterAgent ties every claim to image evidence and lifts report scores 13.5%

    Remote sensing disaster assessment has long stopped at segmentation masks, bounding boxes and category labels, while responders need checkable written reports. DisasterAgent feeds bitemporal imagery to a vision expert toolkit, organizes the tool outputs into a spatiotemporal perception graph, and draws inline-cited evidence from that graph to drive a vision-language model that was never fine-tuned on disasters. On DisasterM3 its average disaster report generation score is 13.5% above the Qwen3.5-27B model serving as its central brain; on RSCC, ST5-SCS is 15.34% above that same base model.

  10. arXiv

    One2N pushes a 0.7° forecast model to 0.07°: four weeks on one 80 GB GPU, tropospheric skill intact

    As Earth-system AI models approach kilometre scales, halving grid spacing quadruples the grid-point count and the end-to-end training cost. One2N splits resolution refinement into successive increments: at each level the complete parent predictor is frozen and only a cross-resolution residual correction module is trained, so the parent's activations and optimizer states are never retained. One refinement level on the NeuralGCM-0.7° baseline yields One2N-0.07°, trained for about four weeks on a single 80 GB A800 GPU; against HRES-0.07° analyses it keeps the parent's tropospheric skill while reducing upper-stratospheric errors and resolving finer structure.