Skip to main content
Search

Microsoft (United States)

22 related articles

Subscribe to updates
  1. bioRxiv

    GENESIS drops the need for control cells: it generates single-cell perturbation responses from experimental covariates alone and nears experimental reproducibility on Tahoe-100M

    Single-cell perturbation atlases cover only a fraction of possible experimental combinations, and existing generative models map measured control cells to perturbed cells without a reference for how good a prediction can be. GENESIS instead generates single-cell expression distributions directly from covariates such as cell line or donor, drug or cytokine, dose and batch, training a lightweight multilayer perceptron with the energy score and needing no control cells at inference. On 530 repeated Tahoe-100M plate pairs its Pearson Δ was 0.430 versus 0.432 for experimental repeats and its energy distance 0.158 versus 0.154, neither significantly different; top-100 DEG overlap was 0.289 versus 0.299, 3.4% short.

  2. arXiv

    History Forcing teaches few-step video models to remember: a 1.3B student reaches 79.82 on 60-second VBench-Long

    Few-step autoregressive video diffusion models gradually lose subject and scene information over long rollouts, and existing distillation targets do not constrain distant history separately. History Forcing builds a context residual from the difference between one autoregressive teacher's predictions with and without extra history, then weights it into the DMD or OPSD distillation target. A 1.3B student reaches a VBench quality score of 85.07 for 15-second generation and a VBench-Long aggregate of 79.82 for 60-second generation, and raises PlayWorld Evolution Quality from 2.35 to 2.70 when the camera leaves and returns.

  3. arXiv

    Software agents no longer get only pass or fail: the policy itself ranks debugging traces within the same outcome

    Reinforcement learning rewards for software engineering agents no longer stop at pass or fail: the current policy compares debugging traces from the same task and verifier outcome, turning those preferences into bounded auxiliary rewards and lifting Qwen3-14B average accuracy on SWE-bench Verified from 32.1% to 34.7%.

  4. arXiv

    Collaboration can backfire when modalities vanish: Lluna keeps a direct per-model prediction and gains 4.26% on average across eight datasets

    Collaboration among pretrained unimodal models can become less accurate than a surviving unimodal model when modalities are missing. Lluna keeps each model's direct prediction as a private path, adds a graph-based common branch as a residual correction, and lets richer availability states guide poorer ones along single-modality removal chains. Across eight datasets in affective computing, scene understanding and medical prediction, it improves Mean-Missing performance by 4.26% on average over the strongest evaluated baseline, with positive collaboration gains at minimal inputs on all eight tasks.

  5. arXiv

    Better task completion, more leakage: 936 executable agent tasks show 32.5% to 94.4% privacy leakage

    AgentCIScope turns privacy cases into 936 executable agent tasks across 11 application environments and finds single-run leakage rates from 32.5% to 94.4% across 12 agents, so strong task completion does not imply contextual integrity.

  6. arXiv

    Code review agents lifted bug catch rate from 47.7% to 71.9% in five months, but weaken on proprietary repositories

    CRBugBench tests code review agents against 1,800 human-fixed bugs: the best configuration catches 79.6% of pre-merge bugs within its top-5 comments, but performance drops noticeably on proprietary industry repositories.

  7. arXiv

    Terminal agents build their own RL environments: up to 25x more valid ones, 75% cheaper each

    AutoEnvScaling turns reinforcement-learning environment design into a terminal task, letting a proposer agent build new environments from the solver's rollouts, raising the valid-environment rate by up to 25 times and cutting cost per valid environment by up to 75%.

  8. arXiv

    Creativity benchmarks are really testing the tail: scoring only a benchmark's rarest items tracks human preference better

    Sorting creativity benchmarks by how rare their items are under a reference corpus makes the tail subset's model ranking agree more closely with LM Arena creative-writing human preferences than the peak subset, so creativity evaluation can be read as task performance in the long tail.

  9. arXiv

    Researchers show mathematically that omitting or truncating importance-weight correction in RLVR implicitly reweights the reward, and propose an SMC correction mechanism

    This work shows mathematically that, in sparse-reward RLVR, omitting or truncating importance-weight correction for enriched rollouts implicitly reweights the defined reward, and decomposes the resulting gradient error into scale, rotation, and variance; the authors develop a sequential Monte Carlo (SMC) weight correction mechanism that, under stability and particle-order assumptions, tempers the exponential compounding of standard correction variance over sample length into additive accumulation, and they apply this analysis to interpret a fine-tune of Qwen3-1.7B on a sparse band of OpenMathReasoning, establishing a concrete mechanism of how enrichment, corrected and uncorrected, helps avoid collapse in sparse domains.

  10. arXiv

    EchoChat recasts empathetic spoken dialogue as a three-stage perception–mental-state–response pipeline, lifting EchoEval perception from 6.96 to 8.23 and reasoning from 6.61 to 8.43

    The work introduces EchoChat, which reformulates empathetic spoken dialogue as a structured cognitive reasoning process spanning perception, mental-state reasoning, and response generation, supported by the EchoDialogue-400K dataset of 400K single-turn samples plus 4K multi-turn sessions, Acoustic-Anchored Attention, and a stage-aware RL objective with Step-Decomposed Credit Assignment, together with the 1K-instance expert-annotated EchoEval benchmark; experiments report gains over 12 compared models with perception 8.23, reasoning 8.43, and response 9.26, and an average Spearman correlation of 0.83 between LLM-based scores and human judgments.