Evolve Quick Reviews
Public Paper Reviews
Structured preliminary reviews of recent open-access papers, generated and continuously updated by the Evolve analysis engine. Each review condenses what the paper claims, how solid the evidence is, and the questions a rigorous peer reviewer would ask — English analysis with abstracts in your language. These are independent AI analyses, not peer review; the canonical record always lives at the publisher.
HoliBench: A Cross-Platform Benchmarking and Deployment Toolkit for Foundation Models in CPS-IoT Applications
Inesh Chakrabarti, Zejun Xiong, Pragya Sharma, Mani Srivastava
Unified FM accuracy-latency-energy profiling across 7 platforms; useful honest findings, but PAL validation is indirect and composability rests on one device.
Understanding Operator Attitudes Toward AI-Supported Decision Making in Maritime Operations
Doreen Jirak, Armeen Saroukanoff, Dirk van Rooy
Survey of 66 seafarers finds positive AI attitudes and stable trust across scenarios; valuable expert data, but null age/experience results are underpowered.
Explaining Time Series Forecasting with Horizon-Resolved Attribution
Seunghan Lee, Jun Seo, Jaehoon Lee, Junhyeok Kang et al.
Proposes HRX, per-step explanations for TS forecasting; extensive 56K-run evidence supports the horizon axis, though authors concede no downstream utility yet.
Learning structural balance of graphs from quantum spectral features
Stefano Scali, Oleksandr Kyriienko
DOS moments of an Ising encoding estimate signed-graph frustration index well at n≤12; quantum extraction only simulated, advantage claimed but untested.
LOCUS: Task-Aware Low-Rank Post-Training for Token-Efficient Language Generation
Dongfang Zhao
Claims low-rank subspace post-training cuts output length up to 40% with the loss unchanged; protocol-matched but utility rests on a near-chance metric.
The widening evaluation gap in medical large language model research 2023 to 2026
Raad Bin Tareaf, Murad Al-Rajab, Samia Loucif
Claims rigorous medical-LLM evidence trails the models it evaluates; a reproducible meta-research effort, with real caveats on hand-set release dates and title/abstract-only detection.
SIRF: A Spec-Internalized Risk Foundation Model for Industrial Content Risk Control
Suwan Wu, Yumeng Lin, Pengcheng Yuan, Xiaolong Jiang
Claims +15.1pp Black Recall@P95 via policy-internalizing CPT on same-source 8B; controlled study credible but proprietary, single-platform, n=1000 eval.
Sparsity Regularized and Robust Mean Variance Portfolio Selection Under Ellipsoidal Uncertainty
Deniz Akkaya, Emre Can Yayla, Buse Şen, Mustafa Ç. Pınar
Robust sparse mean-variance portfolios via ℓ0 + ellipsoidal uncertainty; mostly sound theory, one proof gap; BnB beats Gurobi on time, suboptimal on S&P500.
RAG-Safety-Bench: Reliable Evaluation of Retrieval-Augmented LLM Safety
Adithiyan Rajan Indira Saravanan, Kathleen C. Fraser
A controlled four-condition RAG safety benchmark across five small open LLMs; design is strong, but caveats: n=5 stats, single run, LLM-built data, judges disagree.
Recognizing Is Not Reversing: A Controlled Inversion Test of Fact-Preserving News Framing
Yi Liu
LLMs preserve news facts (FactF1 0.84) yet reverse <8% of injected framing edits despite correct recognition; new benchmark, but data is generator-validated.
Dynamic language model representations for multi-objective reaction optimisation
Joshua W. Sin, David Ming Segura, Bojana Ranković, Siu Lun Chau et al.
Joint LLM-GP text featurisation speeds multi-objective reaction BO; strong benchmarks and two prospective HTE wins, yet no baseline arm ran prospectively.
Component-Aware Differential Privacy for Federated Multilingual Speech-LLMs
Jordi Luque, Fernando López, Aleix Sant
α-split two-pool DP clipping fixes budget collapse in federated speech-LLMs; diagnosis strong, but gradient-inversion claim untested and final ε unreported.
A Unified Per-Token Gating Family for On-Policy Distillation: FKL/RKL Mixing with Multi-Channel and Bias Coefficients
Suwan Wu, Yumeng Lin, Pengcheng Yuan, Xiaolong Jiang
A 4-coefficient per-token FKL/RKL gating family unifying EOPD/ToDi as 1D proxies; honest stats reporting, yet evidence stays directional (single-seed, n=3).
Whisper-Based Speech Transcription from Videos Across Multiple Languages for Cross-Cultural Understanding
Michael Picheny
Claims a Whisper pipeline cuts avg WER from 29.7% to 21.7% across 7 languages; useful practical work, but references are unverified YouTube captions.
Beyond Word Error Rate: A Switch Aware Evaluation of ASR and Audio Language Models on English Yoruba Code-Switched Speech
Chibuzor Okocha, Christan Earl Grant
Claims WER hides code-switch failure: 11-system English–Yoruba benchmark; core dissociation convincing, but single corpus and small-n stats limit generality.
Model-Aware Schedules Improve Generation via Fiberwise Optimal Transport
Luyi Jia, Boyan Zhang, Yilun Liu, Steffen Rulands
Fiberwise-OT model-aware schedules for DDPM/flow matching give consistent FID gains with solid controls; λ tuning and single-seed ImageNet-64 temper the claims.
Predicting Privacy Leakage from Weight Spectral Density
Richard J. Preen, Jim Smith
First empirical link between WW spectral metrics and MIA risk; correlations are suggestive but capacity confounds and MLP-only scope limit generality.
Thinking with Looped Flows
Ayhan Suleymanzade, Chanhyuk Lee, Floor Eijkelboom, Nicholas M. Boffi et al.
Looped flows train stateful denoisers with local flow losses for recurrent reasoning; strong gains over TRM/GRAM on ARC, but baselines use published numbers.
SpecGuard: Inference-Time Backdoor Detection For Free
Rui Wen, Ahmed Salem, Andrew Paverd, Mark Russinovich et al.
Free LLM backdoor detection from speculative-decoding accept/reject traces; broad strong results, but naturalistic payloads evade it and no code is released.
Target leakage, not model class, explains reported accuracy in survey-based cardiovascular screening: a leakage-tiered audit of glass-box and tabular foundation models
Raad Bin Tareaf, Murad Al-Rajab, Samia Loucif, Samer Ellaham et al.
Claims BRFSS MI-screening's ~0.89 AUROC stems from two leaked post-diagnostic features, not model class; rigorous multi-axis audit, convincing but BRFSS-only.
Near-Optimal Reinforcement Learning with Multi-Step Transition Lookahead
Corentin Pla, Hugo Richard, Marc Abeille, Vianney Perchet
Proves look-ahead planning NP-hard for any fixed discount, gives an RPTAS and DLA-UCB with tabular-matching regret; proofs sound, no experiments reported.
From Parameters to Answers: How LLMs Retrieve and Use Their Internal Knowledge
Wenkang Wei, Yuan Fang, Renhe Jiang, Hong Cheng et al.
Claims a layerwise shift from request routing to fact content in LLM answers; careful interventions, but generality rests on one task family, chiefly Qwen.
RetroThinker: Enabling Retrospective Thinking in Speech LLMs
Yi-Jen Shih, Puyuan Peng, Abdelrahman Mohamed, David Harwath
RetroThinker trains streaming SpeechLLMs to self-correct CoT forward-only; +11 accuracy pts on spoken GSM8K at similar latency. Single-benchmark evidence.
IndicTriMix: Developing Language Identification Datasets and Models for Tri-Language Code-Mixing
Pruthwik Mishra, Rudra Trivedi, Avi Patel, Ashok Urlana et al.
First tri-language code-mixed LID benchmark; valuable released data, but Hindi F1 collapses on LLM eval sets (0.34–0.65) and absent baselines weaken claims.