Academic Intelligence · Curated Daily

Explore the Frontier of Global Academia

AcademicHub aggregates real-time literature from top journals and preprint platforms. Build your personal research radar and let large language models compile cross-disciplinary analysis briefings automatically.

01.
medRxiv (Medicine) 2026-06-16

Upper airway disease in primary ciliary dyskinesia: Clinical management and factors influencing decision-making, a multicentre analysis

Background Upper airway disease is common in primary ciliary dyskinesia (PCD), but management evidence is limited. We aimed to describe management practices and identify factors influencing management decisions. Methods Using data from the Ear-Nose-Throat (ENT) Prospective International Cohort of patients with PCD (EPIC-PCD) and an ENT-specialist survey across participating centres, we described management practices recorded at routine follow-up. We assessed clinical factors associated with practices via mixed-effects logistic regression models. In a subgroup of patients, we assessed factors associated with initiation or discontinuation of practices. Results We included 579 patients: median age 15 years, 46% female. Nasal rinsing (54%) and nasal corticosteroids (22%) were most frequently prescribed. Among 466 patients with available data, 47 had grommets (10%) and 42 hearing aids (9%). Nasal corticosteroids and rinsing were more frequently prescribed in patients with polyps (odds ratio [OR] 3.74, 95% confidence interval [CI] 1.80-7.76; OR 3.39, 95% CI 1.37-8.37) or turbinate hypertrophy (OR 1.89, 95% CI 1.03-3.47; OR 2.89, 95% CI 1.55-5.38), and upper airway nebulisation in patients with frequent nasal symptoms (OR 2.86, 95% CI 1.11-7.39). Management practices differed between centres, as seen also by the specialists survey responses. In 177 patients with multiple visits, initiation of nasal rinsing was associated with frequent nasal symptoms (OR 3.18, 95% CI 1.24-8.18) and turbinate hypertrophy (OR 3.21, 95% CI 1.20-8.59). Conclusion Upper airway disease management in PCD varies and is partly guided by symptom burden and clinical findings. This variation across centres highlights the need for care standardisation and PCD-specific management guidelines.

02.
arXiv (quant-ph) 2026-06-12

Metabolic quantum limit to the information capacity of magnetoencephalography

arXiv:2511.06401v3 Announce Type: replace-cross Abstract: Magnetoencephalography measures the magnetic fields generated by neural currents using quantum sensors such as superconducting quantum interference devices and atomic magnetometers. Here we combine the energy resolution limit of magnetic sensing with the metabolic power available to neural currents to derive a technology-independent bound on the information capacity of MEG. The bound factorizes into geometry, metabolism, and Planck's constant, and gives an estimated maximum information rate of 2.2~Mbit/s for representative human-brain parameters. Further, we show that the externally measurable magnetic field has a finite angular bandwidth, with high multipole components being geometrically attenuated and falling below the quantum-limited noise floor. This yields an information-limited spatial scale of order $1~cm$ and renders the accessible measurement space effectively finite-dimensional. The energy resolution limit therefore defines an information-theoretic Nyquist scale for magnetoencephalography, beyond which denser spatial sampling provides redundant measurements rather than additional recoverable information. Since the energy resolution limit also makes the noise variance grow linearly with measurement bandwidth, temporal and spatial bandwidths compete, producing a fundamental spatio-temporal trade-off. These results show how quantum-limited measurements constrain the observable complexity and information content of noninvasive brain imaging, providing a quantitative link between fundamental physics and neuroscience.

03.
arXiv (CS.AI) 2026-06-16

S-SPPO: Semantic-Calibrated Self-Play Preference Optimization

arXiv:2606.01561v2 Announce Type: replace Abstract: Aligning Large Language Models (LLMs) with human preferences is often formulated via Direct Preference Optimization (DPO). However, the standard Bradley-Terry instantiation of DPO is limited in modeling common departures from transitivity in human preferences. To address this, recent work has introduced Self-Play Preference Optimization (SPPO), which iteratively refines the policy by training on self-generated win-lose pairs. Our investigation, however, reveals a critical instability in SPPO: the optimization is prone to policy degeneration when the preference oracle assigns overly confident wins to semantically indistinguishable responses. To mitigate this, we propose S-SPPO, a dual-space semantic calibration framework comprising: i) Supervision Calibration via semantic gating, which anneals win rate targets toward the maximum-entropy baseline as semantic overlap increases; and ii) Representation Calibration via latent repulsion to enforce geometric diversity to prevent manifold collapse and maintain latent diversity between chosen and rejected samples. Theoretically, we show that the calibration preserves the constant-sum game structure, facilitating convergence to a Nash Equilibrium. Empirically, S-SPPO avoids the performance degradation seen in prior methods, achieving 52.19% win rate and 47.46% length-controlled win rate on AlpacaEval 2.0 with Llama-3-8B, without using additional human-annotated preferences during training. The code will be available at https://github.com/xiwenc1/s-sppo.

04.
arXiv (CS.CL) 2026-06-16

The Value Axis: Language Models Encode Whether They're on the Right Track

We investigate whether language models internally track the value of their current trajectory, defined as the likelihood that their ongoing strategy will achieve their goals. Using synthetic, in-context reinforcement learning data, we construct a "value" axis for Qwen3-8B. We find that activations along this axis distinguish between high vs. low verbalized confidence, rollouts without and with backtracking, and correct vs. corrupted code. Steering towards high value causally suppresses self-correction and reduces explanatory verbosity, while steering towards low value induces backtracking and exploration. We demonstrate that direct preference optimization (DPO) can increase the internal value of rewarded behaviors (e.g. use a certain word), causing the model to act more confidently after exhibiting them. Finally, we apply the value axis to study in-the-wild settings. For example, we find that Qwen assigns low value to politically sensitive chat queries after post-training and that supervised fine-tuning increases internal confidence within the training domain. Our results suggest that language models linearly encode an estimate of expected goal success that modulates their confidence in pursuing a direction.

05.
arXiv (CS.AI) 2026-06-15

Closing the Reflection Gap: A Free Calibration Bonus for Agentic RL

Authors:

arXiv:2606.14211v1 Announce Type: new Abstract: LLMs are increasingly deployed as agents that interact with external environments and observe feedback such as execution results, error messages, and tool outputs. A well-functioning agent should be able to leverage this feedback to accurately assess its own performance. Yet we find a persistent reflection gap: LLM agents tend to mis-assess their own outputs after observing concrete environment feedback – even for questions they correctly answered – and standard RL barely helps due to a credit-assignment mismatch. To close this gap, we propose RefGRPO, a simple yet effective fix that augments standard RL algorithms with two key ingredients: a free calibration bonus computed by contrasting the agent's own reflection with the actual outcome (requiring no additional reward model, LLM judge, or external annotation), and a dynamic schedule on its coefficient. Compared to standard RL baselines, our method simultaneously improves reflection calibration (e.g., reduces underconfidence rate $44.4\% \to 7.7\%$) and task accuracy (e.g., $75.1\% \to 76.5\%$) on text-to-SQL across five benchmarks. The resulting calibrated reflection turns the agent into its own verifier grounded in environment feedback, which further enables (i) better self-improvement that uses reflections as pseudo-rewards without outcome supervision, and (ii) more effective test-time selective prediction by committing only to rollouts flagged as correct.

06.
medRxiv (Medicine) 2026-06-17

Performance of five risk stratification tools for paediatric pneumonia against WHO scores using data from the PediCAP trial in sub-Saharan Africa

Background Risk stratification tools for childhood pneumonia have been proposed to improve identification of children at highest risk of death, particularly in low-resource settings. However, their added value over the WHO Integrated Management of Childhood Illness (IMCI) criteria and danger signs remains uncertain. Methods We conducted a secondary analysis of a multi-country randomised controlled trial of children without HIV hospitalised with pneumonia in Mozambique, South Africa, Uganda, Zambia, and Zimbabwe. We evaluated the performance of five published risk scores alongside WHO IMCI severity classification and danger signs. Discrimination for (1) in-hospital mortality, (2) 28-day mortality, and (3) 28-day readmission or death was assessed using area under the receiver operating characteristic curve (AUC). Comparative performance and clinical utility were examined. Results Of the 1010 participants, 18 (1.8%) died in hospital, 22 (2.2%) died in hospital or in the 7 days post-discharge, and 63 (6.2%) died or were readmitted by day 28. Univariate case-fatality rates were highest for variables associated with malnutrition, convulsions, and hypoxaemia. All risk scores demonstrated moderate discrimination for in-hospital and in-hospital+7-day mortality (AUC range approximately 0.75-0.84), with no meaningful differences between models, and performed similarly to the WHO danger signs and IMCI severity classification. In contrast, all approaches performed poorly in predicting 28-day readmission or death (AUC approximately 0.54-0.58). No risk score consistently outperformed simple clinical criteria. Conclusions In this multi-country dataset, we found no evidence that published paediatric pneumonia risk scores meaningfully outperform WHO IMCI-based clinical assessment for predicting mortality. The relatively small number of mortality events limits precision, and modest differences cannot be excluded. These findings suggest that, in low-resource settings, strengthening implementation of existing WHO clinical criteria may be more effective than adopting more complex prediction tools.

07.
medRxiv (Medicine) 2026-06-22

Use of the Pharmacy First service in England in the first 12 months: geographic variation and health system context

Objectives: The Pharmacy First (PF) service was introduced across England from 31 January 2024 to expand the clinical role of community pharmacies and improve access to primary care. This paper describes use of PF in its first 12 months, in terms of uptake, access routes, consultation outcomes, geographic variations, service costs and antimicrobial supply. Methods: A descriptive analysis of all PF consultations submitted for payment to NHS Business Services Authority in England between 31 January 2024 and 31 January 2025. Pharmacy-level consultation data were linked to national data on population, location and pharmacy characteristics. PF use was examined using population-standardised consultation rates and consultations per pharmacy. Results: During the first year of implementation, 2,205,731 PF consultations were recorded as delivered across 11,349 pharmacies, with payment of GBP123 million to pharmacies. Uptake increased steadily over time. Most consultations were for acute sore throat (33%) and uncomplicated urinary tract infection (27%), with corresponding antibiotics, phenoxymethylpenicillin and nitrofurantoin being the most supplied. Most people self-referred (74%) into the service, with 95% of consultations managed without onward referral. Substantial geographic variation was observed. Northern regions had higher use based on the eligible population. The South East and Midlands had higher activity per pharmacy. London showed a distinct pattern, with higher self-referral into the service, lower medication supply and higher referral to other healthcare services. Higher consultation volume was weakly associated with pharmacy characteristics, including opening hours, pharmacy type and retail setting, and local context, in terms of socio-economic and geographic factors. Conclusions: PF had immediate uptake and is operating primarily as a direct-access model for common acute conditions. Findings suggest that PF is contributing to improved access to care and may shift demand away from general practice. However, the service uptake appears to be shaped by geographic location, proximity to other healthcare services and pharmacy characteristics.

08.
arXiv (CS.AI) 2026-06-19

UniMM: A Unified Mixture Model Framework for Multi-Agent Simulation

arXiv:2501.17015v2 Announce Type: replace Abstract: Simulation plays a crucial role in assessing autonomous driving systems, where the generation of realistic multi-agent behaviors is a key aspect. In multi-agent simulation, the primary challenges include behavioral multimodality and closed-loop distributional shifts. In this study, we formulate a unified mixture model (UniMM) framework for generating multimodal agent behaviors, which can cover the mainstream methods including regression-based mixture models and discrete NTP models. Furthermore, we introduce a closed-loop sample generation approach tailored for mixture models to mitigate distributional shifts. Within the UniMM framework, we recognize critical configurations from both the model and data perspectives. We conduct a systematic examination of various model configurations, and comprehensively characterize their effects. Moreover, our investigation into the data configuration highlights the pivotal role of closed-loop samples in achieving realistic simulations. To extend the benefits of closed-loop samples across a broader range of mixture models, we further introduce a temporal disentanglement-and-alignment mechanism to address the shortcut learning and off-policy learning issues. Leveraging insights from our exploration, the distinct variants proposed within the UniMM framework, including discrete, anchor-free, and anchor-based models, all achieve state-of-the-art performance on the WOSAC benchmark.

09.
arXiv (CS.CV) 2026-06-18

Characterizing Brazilian Atlantic Forest Restoration Outcomes with Geospatial AlphaEarth Embeddings

Authors:

The Atlantic Forest in Brazil is a critical biodiversity hotspot, yet less than 12-15% of its original cover remains. Although monitoring forest restoration on a large scale is essential, traditional methods are limited by the impracticality of on-the-ground reporting on such a scale and by the saturation of remote-sensing indices such as NDVI. Furthermore, reforestation is a gradual process as opposed to the rapid spectral changes caused by deforestation. In this study, we examine 1,729 restoration sites in S\~ao Paulo, using satellite embeddings from the AlphaEarth Foundation's model to evaluate their effectiveness in characterising early restoration success. We introduce the concept of a 'Reference Trajectory Embedding', defining a metric of restoration success based on cosine similarity to reference sites of mature secondary forest. We observe distinct clusters in embedding space according to different land use and land cover (LULC) types, and we can identify sites with clear change vectors. However, the signal can be noisy, and embeddings may require further fine-tuning to capture and predict site metadata beyond LULC.

10.
arXiv (CS.CL) 2026-06-18

RegMix-D: Dynamic Data Mixing via Proxy Training Trajectories

Data mixture selection is critical for Large Language Model pretraining. Existing methods such as RegMix select a single static mixture by fitting a regression model on small-scale proxy runs. We propose RegMix-D, a simple extension of RegMix to dynamic mixing. Our key observation is that proxy runs produce not only endpoint losses, but also full loss trajectories, which can be used to further improve data mixture. By training regression model on these trajectories, we can predict optimal mixtures at multiple training stages. RegMix-D supports two deployment modes: an offline variant that generates a complete mixture schedule before target training, and an online variant that adapts the mixture during training using observed loss. Experiments on 25B tokens of the Pile dataset with a 1B parameter target model show that RegMix-D consistently improves over RegMix and DoReMi across 13 downstream tasks while remaining proxy-efficient: it surpasses RegMix even with only 128 proxy models (25% of RegMix's proxy compute budget).

11.
arXiv (CS.CV) 2026-06-15

SED:Lightweight Saliency prediction for Event-based data via Distillation

Event-based saliency prediction has gained attention recently, as combining event cameras with saliency estimation can act as an upstream stage that naturally improves the efficiency of downstream eventbased perception at the edge. However, current approaches are either neuromorphic, underperforming on event-based saliency benchmarks, or too heavy for resource-constrained edge applications due to their reliance on transformers or 3D convolutions. Drawing inspiration from efficient convolutional modules, SED and aiming to exploit the temporal information in event data, we propose a lightweight network, trained through knowledge distillation, built on a Depthwise Spatio-Temporal Block (DSTconv) – a factorization of the 3D depthwise separable convolution. Relative to its teacher, our model reduces the model size from 180 MB to 0.32 MB (562x) and the parameter count from 45M to 81k (554x), while matching or outperforming it on the N-DHF1K and N-UCF Sports datasets. Moreover, it generalizes strongly beyond its training distribution, transferring from synthetic to real event data where a model trained from scratch fails.

12.
medRxiv (Medicine) 2026-06-22

Population-Scale, Genotype-First Characterization of Monogenic Diabetes in 374,973 Multi-Ancestry Individuals from the All of Us Research Program

OBJECTIVE To characterize the prevalence and penetrance of maturity-onset diabetes of the young (MODY) in a multi-ancestry population using a genotype-first design. RESEARCH DESIGN AND METHODS We analyzed whole-genome sequencing and clinical data from 374,973 unrelated All of Us participants (42.0% non-European ancestry). We identified pathogenic or likely pathogenic (P/LP) variants in 10 established MODY genes and assessed carrier prevalence, diabetes penetrance, and glycemic profiles. We evaluated age-dependent diabetes risk by comparing carriers with non-carriers stratified by type 2 diabetes polygenic risk score (T2D PRS). RESULTS We identified 370 carriers of P/LP MODY gene variants (0.099%; 1 in 1,013), with similar carrier prevalence among European- and African-ancestry participants (0.105% in both groups). Diabetes penetrance was incomplete (13.4% by age 40; 43.5% by age 60) and varied by etiology: highest for GCK (56.0% by age 60), intermediate for HNF genes (HNF1A/HNF1B/HNF4A; 45.4%), and lowest for non-GCK/HNF genes (ABCC8/INS/KCNJ11/NEUROD1/PDX1/RFX6; 29.0%). In multivariable Cox models using non-carriers in the middle 80% of the T2D PRS as the reference, non-GCK/HNF gene variant carriers had modestly increased diabetes risk (HR, 1.57), similar to non-carriers in the top 10% of T2D PRS (HR, 1.64). These associations were observed in both European- and non-European-ancestry individuals. HbA1c profiles differed by etiology, with stable mild hyperglycemia in GCK variant carriers and greater variability among HNF and non-GCK/HNF gene variant carriers. CONCLUSIONS MODY gene variants showed incomplete, etiology-dependent penetrance across ancestries. Carriers of P/LP variants in lower-penetrance genes had diabetes risk comparable to that of non-carriers with high polygenic susceptibility.

13.
arXiv (CS.AI) 2026-06-11

OmniBioTwin: A System-of-Twinned-Systems Framework for Health Digital Twins

arXiv:2606.11264v1 Announce Type: cross Abstract: Health digital twins (HDTs) promise patient-specific modeling and decision support but current approaches remain structurally fragmented: monolithic models that address a single organ or task lack cross-scale fidelity, while system-level twins lack generalizable architectural frameworks. We propose OmniBioTwin, a System-of-Twinned-Systems (SoTS) framework that organizes HDTs as modular computational entities coupled through explicit interaction operators within a multi-layer network architecture. The framework comprises seven coordinated layers - spanning data integration, autonomous twin modeling, cross-scale coupling, temporal synchronization, and human-in-the-loop decision support. We demonstrate OmniBioTwin by instantiating a multiscale twin for glucagon-like peptide-1 (GLP-1) signaling pathways in Alzheimer's disease, illustrating how molecular, cellular, and organ-level twins can be composed and coupled within a unified system.

14.
arXiv (CS.AI) 2026-06-16

LabOSBench: Benchmarking Computer Use Agents for Scientific Instrument Control

arXiv:2606.16802v1 Announce Type: new Abstract: Current computer-use benchmarks primarily focus on software operation tasks in virtualized systems, whereas scientific instrumentation scenarios require coordinated control over complex interfaces, and feedback-driven parameter adjustment. However, directly evaluating agents on physical high-precision instruments is impractical due to high cost, safety risks, limited accessibility, and difficulty in ensuring reproducible evaluation. This motivates the need for a simulated yet realistic testbed that preserves the operational challenges of scientific instruments while enabling scalable and safe benchmarking. To this end, we introduce LabOSBench, a challenging benchmark for multimodal GUI agents built on a suite of web-based scientific-instrument simulators. Operating directly via a browser, LabOSBench avoids resource-heavy OS virtualization while supporting flexible task configuration and execution-based evaluation. Specifically, LabOSBench constructs 96 subtasks across eight instrument simulators, covering workflows from sample loading, alignment, parameter tuning, and data acquisition to result inspection. We evaluate general-purpose vision-language models, specialized GUI agent models, and advanced agentic frameworks at both subtask and end-to-end levels. Our experiments reveal that while existing agents can complete many structured GUI subtasks, they still struggle with feedback-driven operations and long-horizon workflow execution. Overall, LabOSBench provides a reproducible, low-cost testbed for advancing computer-using agents toward scientific-instrument control.

15.
arXiv (CS.CV) 2026-06-17

High-Fidelity 3D Geometric Reconstruction of Pelvic Organs from MRI: A Hybrid Deep Learning and Iterative Optimization Approach

Patient-specific 3D reconstruction of pelvic organ geometry from MRI is important for pelvic floor modeling and downstream patient-specific analysis. However, while previous studies have focused primarily on either image segmentation or downstream use of 3D models, the reconstruction of high-fidelity, high-quality geometries remains labor-intensive and poorly standardized. The study introduced a hybrid deformable shape modeling framework that integrates deep learning prediction with iterative optimization for the reconstruction of the bladder, uterus, and rectum. The framework consists of three core components: a geometry-aware multi-level deep learning architecture that preserves topological consistency of pelvic organs; a two-stage amortized optimization training strategy that balances global shape capture and local surface refinement; and a holistic synergy mechanism–where iterative optimization provides supervision for deep learning during the training phase, and during inference, deep learning rapidly predicts the global organ morphology, followed by iterative optimization to refine local surfaces and mesh quality. This framework demonstrated marked superiority in geometric fidelity than current mainstream deep learning-based organ reconstruction models. For individual anatomical structures, the reconstructed 3D geometries for the bladder, rectum, and uterus achieved significantly lower Chamfer Distance values and higher Dice Similarity Coefficient scores. In addition, while maintaining high computational efficiency, the proposed architecture yielded superior overall volumetric mesh quality. At the patient level, the framework achieved higher mean values for the 10 worst elements for both minSICN and minSIGE compared to traditional geometric post-processing algorithms.

16.
arXiv (CS.AI) 2026-06-16

E-mem: Multi-agent based Episodic Context Reconstruction for LLM Agent Memory

arXiv:2601.21714v5 Announce Type: replace Abstract: The evolution of Large Language Model (LLM) agents towards System~2 reasoning, characterized by deliberative, high-precision problem-solving, requires maintaining rigorous logical integrity over extended horizons. However, prevalent memory preprocessing paradigms suffer from destructive de-contextualization. By compressing complex sequential dependencies into pre-defined structures (e.g., embeddings or graphs), these methods sever the contextual integrity essential for deep reasoning. To address this, we propose E-mem, a framework shifting from Memory Preprocessing to Episodic Context Reconstruction. Inspired by biological engrams, E-mem employs a heterogeneous hierarchical architecture where multiple assistant agents maintain uncompressed memory contexts, while a central master agent orchestrates global planning. Unlike passive retrieval, our mechanism empowers assistants to locally reason within activated segments, extracting context-aware evidence before aggregation. Evaluations on the LoCoMo benchmark demonstrate that E-mem achieves over 54\% F1, surpassing the state-of-the-art GAM by 7.75\%, while reducing token cost by over 70\%.

18.
arXiv (CS.AI) 2026-06-17

Unlocking LLM Code Correction with Iterative Feedback Loops

arXiv:2606.17514v1 Announce Type: cross Abstract: Large Language Models have shown remarkable capabilities in code generation. However, most existing evaluations focus only on single-attempt accuracy and overlook the iterative refinement process that is central to real-world programming. This study presents a systematic investigation of LLMs' ability to rectify their own code through execution feedback. Using real-world programming problems across four models and two major programming languages, this study evaluates performance using iterative refinement framework where LLMs receive compiler error messages and testcase feedback after each attempt. This study introduces metrics to evaluate code failures, analyze rectification patterns, and compare the effectiveness of reasoning and non-reasoning models, offering actionable insights into both the understanding and practical application of feedback loops in LLM-driven code generation systems. Results show that reasoning models consistently improve over iterations, substantially outperforming non-reasoning models in leveraging feedback, while syntactic and runtime errors are far more tractable than logical or algorithmic failures.

19.
arXiv (CS.CL) 2026-06-18

UMA-Split: unimodal aggregation for both English and Mandarin non-autoregressive speech recognition

This paper proposes a unimodal aggregation (UMA) based nonautoregressive model for both English and Mandarin speech recognition. The original UMA explicitly segments and aggregates acoustic frames (with unimodal weights that first monotonically increase and then decrease) of the same text token to learn better representations than regular connectionist temporal classification (CTC). However, it only works well in Mandarin. It struggles with other languages, such as English, for which a single syllable may be tokenized into multiple fine-grained tokens, or a token spans fewer than 3 acoustic frames and fails to form unimodal weights. To address this problem, we propose allowing each UMA-aggregated frame map to multiple tokens, via a simple split module that generates two tokens from each aggregated frame before computing the CTC loss.

20.
arXiv (CS.AI) 2026-06-17

FinAcumen: Financial Multimodal Reasoning via Self-Evolving Experience Memory Harness

arXiv:2606.17642v1 Announce Type: new Abstract: Financial multimodal reasoning requires agents to coordinate numerical computation, retrieval, visual interpretation, and temporal grounding across heterogeneous evidence sources. Existing tool-augmented agents improve execution fidelity, yet remain largely stateless across episodes, repeatedly rediscovering reasoning strategies and failure patterns. In high-stakes financial settings, this leads to unreliable tool routing, noisy retrieval, and hallucination-prone reasoning. We present FinAcumen, a financial reasoning agent framework centered on selective experience memory for tool-augmented multimodal reasoning. FinAcumen accumulates financially grounded reasoning experience from prior trajectories, distilling successful strategies and failure-derived cautionary rules into a persistent memory bank. During inference, retrieved experiences condition reasoning only when semantic relevance exceeds a calibrated threshold, while irrelevant memory is explicitly suppressed through a fallback mechanism. A deterministic financial tool environment further grounds numerical computation, retrieval, visual decoding, and answer verification.Across four financial multimodal reasoning benchmarks, FinAcumen consistently improves a frozen 8B vision-language model over finance-specialized models and approaches leading proprietary general-purpose models. Further analysis shows that selective experience activation improves reasoning reliability under retrieval uncertainty. Our code is anonymously available at https://anonymous.4open.science/r/FinAcumen

21.
arXiv (CS.AI) 2026-06-19

StaminaBench: Stress-Testing Coding Agents over 100 Interaction Turns

arXiv:2606.19613v1 Announce Type: cross Abstract: We introduce StaminaBench, a benchmark that measures the stamina of coding agents: how many consecutive interaction turns (change requests) they can handle before failing. Unlike the prevailing fraction-of-tasks-solved metric, this matches real vibe-coding where sessions run dozens or hundreds of turns. In StaminaBench, agents implement a REST API server and modify it across a tunable number of procedurally generated follow-up change requests - 100 in our experiments, resulting in codebases of up to 6,000 lines. Tests are generated fully programmatically without LLM involvement, ensuring reproducibility and reliability; change sequences are drawn from either a hardcoded or LLM-driven sampler, both constrained to a structured action space to ensure changes are valid. The agent and the server run in an isolated environment and communicate with the benchmark through HTTP, making testing fully black-box and language-agnostic. We evaluate six agent harnesses paired with seven open-source LLMs across 20 scenarios of 100 turns each and find that: (1) all the tested models fail within 5-6 turns, confirming that vibe-coding-style programming without thorough testing produces bugs; (2) passing test feedback back to the agent and allowing it to retry improves passed turn count by up to 12x; and (3) a good harness is required for strong performance: stronger models exhibit up to a 6x gap between their best and worst harness, while weaker models fail with any harness. We release the benchmark and the generated tasks to enable further research into multi-turn coding agent behavior. Benchmark code and data: github.com/amazon-science/StaminaBench.

22.
arXiv (CS.AI) 2026-06-11

From Awareness to Action: Understanding and Overcoming the Research-Practice Gap in Algorithmic Fairness for Public Health

arXiv:2606.11214v1 Announce Type: cross Abstract: Algorithmic fairness is essential for responsible ML-driven public health research, yet its practical implementation remains limited. To investigate this awareness-action gap, we conducted a sequential mixed-methods study comprising expert interviews, an online survey, and systematic mapping. The expert interviews informed the design of the survey, which in turn revealed fragmented definitions of fairness, limited training and guidance, reliance on external sources, and rare use of formal assessment, mitigation, or monitoring. These findings were subsequently mapped onto three established research-practice gap lenses: the Knowledge-Practice Gap, the Knowledge-to-Action Cycle, and the Knowing-Doing Gap, each offering complementary perspectives. Building on this synthesis, we introduce the Fairness-to-Action framework, which integrates methodological, organizational, and systemic dimensions to identify where translation of algorithmic fairness knowledge stalls. Our analysis shows that fairness remains weakly institutionalized, translation mechanisms are externally driven, and system-level priorities continue to emphasize accuracy over fairness. These insights suggest critical leverage points for advancing safe, fair, and ethical ML-driven public health research practice.

23.
arXiv (CS.CL) 2026-06-17

LLM Features Can Hurt GNNs: Concatenation Interference on Homophilous Graph Benchmarks

Adding LLM-generated node features to graph neural networks (GNNs) is widely reported to improve accuracy on standard benchmarks. We document a contrasting observation: when LLM features are introduced through pure input concatenation (rather than joint training, distillation, or prompt-conditioning), they can systematically degrade accuracy on the same homophilous benchmarks where end-to-end LLM pipelines succeed. With an MLP backbone on the Planetoid public split and bag-of-words original features, concatenating SBERT-encoded GPT-4o-mini TAPE features reduces PubMed test accuracy by -17.0 +/- 0.3 pp and Cora by -4.3 +/- 0.6 pp (CiteSeer -0.6 +/- 0.8 pp, within seed noise). The drop attenuates as we relax each condition (GCN / GCNII / GAT backbones, random splits, smaller encoders) and reverses on medium-homophily WikiCS (+4.4 pp) and ogbn-arxiv (+11.7 pp). To predict when concatenation helps versus hurts, we report a simple measure of LLM-alone discriminability, Delta_sig. Across 9 datasets Delta_sig correlates with the concatenation cost more strongly than homophily at point estimate (r^2 = 0.38 vs. 0.06; N=9, bootstrap CIs overlap). The bootstrap-best change-point is tau = 13.8 pp, and the rule "Delta_sig

24.
arXiv (CS.AI) 2026-06-19

Rethinking Shrinkage Bias in LLM FP4 Pretraining: Geometric Origin, Systemic Impact, and UFP4 Recipe

arXiv:2606.20381v1 Announce Type: new Abstract: FP4 training promises substantial reductions in memory and computation cost for LLM pretraining, yet current FP4 hardware paths and recipes, including NVIDIA Blackwell/Rubin-class systems and AMD MI350-series GPUs, remain centered on E2M1 data elements. In this study, we identify a fundamental limitation of that choice: non-uniform formats such as E2M1 inherently suffer from Shrinkage Bias, a systematic negative rounding error caused by the geometric asymmetry of their representable bins. We show that this bias accumulates multiplicatively across layers and is amplified by the Random Hadamard Transform (RHT), providing a unified explanation for the training instability observed in existing E2M1-based FP4 recipes. In contrast, uniform grids (E1M2/INT4) bypass this grid-geometry error and better convert the improved bucket utilization from RHT into higher quantization quality. Based on this finding, we propose UFP4, a uniform 4-bit training recipe that applies RHT to all three training GEMMs while restricting stochastic rounding to dY alone. On Dense 1.5B, MoE 7.9B, and MoE 124B long-run pretraining, UFP4 consistently achieves lower BF16-relative loss degradation than strong E2M1-based baselines, supported by scaling-law analysis and ablation studies. Our results suggest that future accelerators should support E1M2/INT4-style uniform 4-bit grids as first-class training primitives alongside E2M1.

25.
arXiv (CS.LG) 2026-06-19

Learning to Emulate Chaos: Adversarial Optimal Transport Regularization

arXiv:2604.21097v2 Announce Type: replace-cross Abstract: Chaos arises in many complex dynamical systems, from weather to power grids, but is difficult to accurately model with data-driven methods such as machine learning emulators. While emulators are promising tools for accelerating simulations and solving inverse problems, they still struggle to learn chaotic dynamics, where sensitivity to initial conditions renders exact long-term forecasts infeasible, especially given noisy data. Recent work instead trains emulators to match the statistical properties of chaotic attractors, but these approaches often rely on handcrafted summary statistics or large, diverse multi-environment datasets. In this work, we propose a family of adversarial optimal transport objectives that can jointly learn high-quality summary statistics and a physically consistent emulator from a single noisy trajectory. We theoretically analyze and experimentally validate a Sinkhorn divergence formulation (2-Wasserstein) and a WGAN-style dual formulation (1-Wasserstein) of our approach. Numerical experiments across a variety of chaotic systems, including ones with high-dimensional spatiotemporal chaos, show that emulators trained using our proposed objectives have significantly improved long-term statistical fidelity.