Skip to main content
Search

Arizona State University

9 related articles

Subscribe to updates
  1. arXiv

    Agents must now identify the PDE themselves: RSI-PDEbench tests them on 1,001 cases, lifting mean pass rates by up to 181.9% across four models

    Recovering a partial differential equation from sampled solution data demands expert work in equation discovery, coefficient estimation and numerical solver configuration. RSI-PDEbench turns that workflow into a benchmark of 1,001 cases over 42 equation templates, split into noise-free, 50% Gaussian-noise and unseen-template-with-shifted-coefficient settings; the accompanying RSI-PDE agent shortlists candidate equations by residual, refines coefficients and solvers with relative feedback, and keeps improvements behind a validation gate. Across four frontier models its mean pass rate rises 106.5–181.9% over single-run baselines and 3.0–19.6% over native agent harnesses.

  2. bioRxiv

    Guidelines move on, answers don't: with a strong model, retrieving several ASCO guideline versions made stale answers about four times as frequent as answering without retrieval

    Oncology guidelines are revised as evidence emerges, so an answer can faithfully cite a real guideline and still cite a superseded version. ASCOBench ties 288 questions to specific versions of ASCO breast and prostate cancer guidelines, with oncologist-reviewed reference answers; about one in four version-sensitive drafts needed revision or rewriting, against fewer than one in ten factual ones. Only 1 of 14 superseded guideline documents states that it has been replaced, and for the strong model Gemini retrieval over several guideline versions made stale answers about four times as frequent as answering without retrieval, while the verification-first SentryLine lowered incorrect answers to 4.2-5.2%.

  3. arXiv

    LinkedIn and university collaborators find that 7B-to-72B students trained on rejects from smaller frozen models do better on code and math while using less compute

    Holding prompts, chosen responses, SeqKD initialization, DPO objective, and training budget fixed while varying only the reject source, the study finds that across Qwen2.5 students from 7B to 72B, rejects generated by smaller frozen Base models with less inference compute train stronger students than the students' own rejects on code generation and mathematical reasoning, and it derives a finite-horizon utility bound in a linearized feature model that motivates three interventions: source mixing, prompt reassignment with lexical permutation, and reference-likelihood reselection.

  4. arXiv

    A survey unifies joint, cross-modal and joint-editing video-audio generation on one distribution and gives the first systematic taxonomy of joint audio-visual editing: nine categories, 28 edit types

    Observing that video and audio are perceived together yet most generative models treat them in isolation, this survey casts joint generation, cross-modal generation (one modality from the other) and joint editing as three problems defined on a single distribution over audio-visual pairs, organizes the field around one question—how the output is kept coherent across modalities in time and semantics—proposes a taxonomy along five design axes (generation strategy, audio representation, video representation, alignment enforcement, pretraining reuse), and, to the authors' knowledge, provides the first systematic taxonomy of joint audio-visual editing, mapped as nine edit categories spanning 28 edit types, together with methods, datasets and metrics for each setting and a closing list of open pr

  5. arXiv

    CUA-SWE tested 105 executable tasks and found that giving agents the running interface lifted GPT-6-Astra's four-domain mean from 11.3% to 59.9%

    The authors introduce CUA-SWE, a benchmark, interactive development environment, and evaluation pipeline spanning 105 tasks across Web, Game, DevOps, and Mobile, in which an agent edits code, runs commands, operates the running software, and inspects screenshots within a single task, with task-specific deterministic tests deciding the outcome; comparing code-only and hybrid CUA conditions across frontier models, GPT-6-Astra reaches the highest four-domain mean of 59.9%, and every frontier model in the four-domain comparison achieves higher aggregate success with hybrid access, with gains concentrated on tasks whose requirements must be recovered from application materials such as drawings, service contracts, or graphical reference cards.

  6. GateScope black-box audits 10 commercial LLM API gateways: some identify gpt-5 as the claimed model only 13.09% of the time, and o*ey bills 62.8% above expected

    The authors introduce GateScope, a lightweight black-box auditing framework that uses only public APIs to evaluate LLM API gateways along response content, multi-turn conversation consistency, billing accuracy, and latency characteristics; controlled validation on official endpoints yields an average F1 of 0.968±0.085 across 24 models, and auditing 10 commercial gateways reveals identification rates as low as 13.09% for gpt-5, degraded multi-turn memory checkpoints, a 62.8% billing gap for o*ey on gpt-4o, and markedly higher latency variation for b*ie.

  7. arXiv

    MeshHeal uses two-timescale peer review to reach 0.839 degraded-phase accuracy at 51k tokens per task on BBH, MATH, and MMLU-Pro, versus Symphony's 0.807 at 115k

    The work introduces MeshHeal, a fully decentralized self-healing framework that couples ability-matched peer review across two timescales to address gray failures in decentralized LLM-based multi-agent systems: at the fast timescale, an adaptive hierarchy escalates uncertain or low-scoring outputs from repeated single-reviewer evaluation to committee deliberation and, when needed, correction before use; at the slow timescale, a task- and ability-conditioned peer-relative detector aggregates scores to distinguish persistent degradation from ordinary output variation, trigger mandatory committee review, and eventually exclude degraded agents from ordinary routing, with recovery probes providing fresh evidence for reintegration; the authors also introduce Model-Backed MAS Evaluation, which ti

  8. arXiv

    Across 25 models, a shared L0 attention routing circuit marks VQ-tokenized VLMs, and ablating it alone cuts open-ended object hallucination by 31% relative

    Using activation patching across twenty-five models spanning eight LLM families, the work identifies an early-layer (L0) attention routing circuit shared by VQ-tokenized vision-language models, proposes a three-gate diagnostic that isolates ten models carrying it and rejects the other fifteen, shows via a single-variable architectural swap (LLaVA-1.6 CLIP+MLP to VQ+Linear) that vector quantization is the source of the pathological signal, and finds that only L0 ablation reduces object hallucination in open-ended generation (CHAIR_i down 31% relative) while tuned DoLA and VCD do not.

  9. Spatial Configuration of Residual Pancreatic Cancer Is Associated with Recurrence Risk

    In 203 patients with pancreatic ductal adenocarcinoma who received neoadjuvant therapy and curative-intent resection and were restricted to minor pathologic response, an AI-enabled pipeline segmented cancer and stroma from routine H&E whole-slide images and quantified spatial composition and configuration, finding that a fragmented, interface-rich tumor-stroma ecology was independently associated with shorter disease-free survival; two spatial risk models (cancer mean shape index plus stromal shape index variability, adjusted HR 1.71, P = 0.003; mean stromal patch area plus edge density, adjusted HR 2.19, P = 0.

You’re all caught up