Skip to main content
Search

Chinese University of Hong Kong

35 related articles

Subscribe to updates
  1. arXiv

    Robust Q-learning drops the generative model: one Markovian trajectory suffices, with near-minimax sample complexity

    For robust Markov decision processes with total-variation uncertainty sets, a model-free asynchronous Q-learning algorithm relies only on a single Markovian trajectory from a behavior policy, without a generative model or queries to arbitrary state-action pairs. It estimates robust Bellman updates through a resampled surrogate next state, uses $|S||A|$ space, and reaches entrywise ϵ\epsilon accuracy with a sample complexity of order ∣S∣∣A∣tmixϵ2μmin⁡\frac{|S||A| t_{\mathrm{mix}}}{\epsilon^2 \mu_{\min}} up to logarithmic factors, carrying the extra factor inherent to Q-learning.

  2. arXiv

    Enzyme–reaction pairing moves beyond sequence similarity: RIVER pools evidence with multiple agents and places all four target enzymes near the top of 667 candidates

    RIVER is a training-free multi-agent retrieval framework that organizes large language models, biochemical computational tools and external knowledge resources into traceable evidence chains for bidirectional enzyme-to-reaction and reaction-to-enzyme retrieval. In four experimentally supported enzyme–reaction cases drawn from an E. coli metabolomics study, it ranked among 667 candidate proteins with a mean reciprocal rank of 0.833 using Qwen3.6-35B-A3B and 0.750 using GPT-5.6 Luna, above the general-purpose coding-agent combinations and domain methods listed in the comparison table. On public benchmarks it reached Top-1 query success rates of 82.3% on Enzyme-405 and 72.8% on Orphan-335.

  3. arXiv

    SurgTAP-3D turns 520K uncalibrated surgical video frames into 1.4M 3D point trajectories

    Vision-language-action models and world models for minimally invasive surgery need large-scale 3D supervision, yet public surgical datasets are largely ex-vivo or simulated and annotate either tissue or instruments, leaving tool-tissue contact regions unmarked. SurgTAP-3D converts uncalibrated monocular surgical videos into pseudo-scale camera poses, depth and 2D/3D point trajectories through contact-region prompting, active-tool segmentation, dynamic bundle adjustment and global 4D refinement, gathering 4,549 action clips, 519,642 frames and about 1.38 million trajectories across robotic and manual procedures. On the SurgT, SurgPose and STIR benchmarks, the pipeline reaches an average AJ3D of 56.57 for 3D tracking, above the compared methods in the table.

  4. arXiv

    A small model plays skeptic to a large one: DoubtLess trims Qwen3-32B's chain of thought by 31% without retraining

    Reasoning models often keep re-checking after they have reached an answer, stretching the chain of thought. DoubtLess has a small model that shares the large model's tokenizer read the same chain twice, once told to go straight to the answer and once told to double-check every step, and adds the scaled difference of the two logit vectors to the large model's logits at every step, acting only at decoding time with no weight changes. On seven competition-mathematics, graduate-science and programming benchmarks, Qwen3-32B steered by Qwen3-0.6B uses 31% fewer thinking tokens at unchanged accuracy, a larger cut than five inference-time baselines.

  5. DeepDelivery mines the human proteome for mRNA delivery scaffolds: 512 candidates, and 12 of 15 tested form particles that deliver

    DeepDelivery treats the human proteome as a candidate pool for mRNA delivery scaffolds: deep learning models trained on viral capsid proteins, filtered by a dual-model screen, nominated 512 candidates from 109,375 human sequences. Of 15 candidates tested experimentally, 12 formed mRNA-encapsulating particles and mediated functional delivery; RTL8B reached 73.2% transduction in HEK293T reporter cells, above the reference proteins Arc (56.5%), PEG10 (53.1%) and PNMA2 (67.4%). The results support human proteins as a source of modularly engineerable delivery scaffolds, while immunogenicity and in vivo behavior remain untested.

  6. aclanthology.org

    RNA inverse folding stops reading only the backbone: treating molecular surface as a design constraint tops sequence recovery on all three benchmarks

    RNA inverse folding recovers a nucleotide sequence from a target 3D structure, and earlier methods leaned on secondary structure and backbone topology while rarely constraining the molecular surface that shapes intermolecular interactions. gRFold fuses a surface point cloud, a backbone graph and secondary structure through hierarchical cross-attention to drive autoregressive decoding, then re-ranks candidates at inference by agreement between predicted folds and the target base-pairing pattern. On the RNA-protein complex, monomeric RNA and RNA-Puzzles benchmarks, gRFold reaches sequence recovery of 49.65%, 51.67% and 45.00%, the highest among the compared methods, and ablations show the surface and backbone branches contribute complementary gains.

  7. arXiv

    Agents must ask real clients what they left unsaid: across 49 projects, the best delivery score is 0.5510

    FDE-Bench turns 49 real customer projects into end-to-end delivery tasks where agents must elicit unstated requirements from a simulated customer and iterate, and across 18 agent configurations the highest mean delivery score is only 0.5510 against an FDE reference of 1.0.

  8. arXiv

    Scientific agents stall on judgment: four of five frontier models pick the plan a senior researcher overruled

    A new benchmark turns senior researchers' corrections of scientific AI agents into 248 two-way judgment questions: given only the project record from before the correction, four of five frontier models more often chose the original plan that a human had overruled, topping out at 63.5% against 73.4% for fresh researchers. The cases come from 26 real projects across six fields run over roughly half a year, and each asks a model to choose between the original plan and the human-corrected alternative.

  9. arXiv

    Small-model reasoning distillation drops the fixed question set: 1,000 trajectories lift a 4B model from 62.9 to 75.3 on IMO-AnswerBench

    Long-chain reasoning distillation no longer fixes its training questions up front: BoundaryEvo repeatedly has the student and a fixed expert answer the same candidates, keeping only questions the expert solves reliably and the student has not mastered, and with 1,000 training examples raises Qwen3.5-4B accuracy on IMO-AnswerBench from 62.9 to 75.3.

  10. arXiv

    A single agent only reweights SFT data: splitting execution, four analyses and synthesis lifts IFEval to 63.33 by round six

    Language model agents are increasingly handed the post-training stack, yet a single agent converges on supervised fine-tuning and spends its effort reweighting data categories. AutoSFT splits the loop into an executor, four analysis subagents that read each round from training data, failures, training signals and history, and a synthesis subagent that writes new examples for undersupplied, verifier-checkable capabilities. Every round retrains from the same base checkpoint, so each component is attributable: analysis reaches 63.33 on IFEval at round six, above the single agent's best of 62.35 across 50 rounds.