Skip to content

Research

Page 87 of 166

OSGuard — A Benchmark for Safety in Computer-Use Agents

research note

OSGuard — A Benchmark for Safety in Computer-Use Agents

·8 min read·Mina Mohammadmirzaei, Jeffrey Flanigan

OSGuard addresses a critical blind spot in evaluating computer-use agents — while many benchmarks assess whether agents complete tasks as instructed, they often ignore safety failures where unsafe s…

researchcomputer-use-agentssafety-benchmarkaction-level-oversightrisk-augmentation

Read note → Source paper ↗

Asymptotically Optimal Codes for Correcting Burst Deletions and Insertions in Labeled DNA Sequences

research note

Asymptotically Optimal Codes for Correcting Burst Deletions and Insertions in Labeled DNA Sequences

·8 min read·Wenhao Liu, Zhengyi Jiang, Zhongyi Huang et al.

This paper addresses the challenge of correcting burst synchronization errors—specifically bursts of consecutive insertions or deletions—in labeling sequences used in DNA-based data storage

researchdna-data-storageburst-error-correctionrun-length-limited-codessynchronization-codes

Read note → Source paper ↗

CANN-EUCLID — unsupervised constitutive artificial neural network model discovery from full-field data

research note

CANN-EUCLID — unsupervised constitutive artificial neural network model discovery from full-field data

·8 min read·Benjamin Alheit, Siddhant Kumar, Mathias Peirlinck

This paper addresses the challenge of discovering interpretable, nonlinear constitutive material models directly from full-field displacement and reaction force data without requiring local stress …

researchconstitutive-model-discoveryunsupervised-learningneural-networkshyperelasticity

Read note → Source paper ↗

CARE — Controlling LLM-Generated Policies through Auditable Review of Evidence in Scientific Experimentation

research note

CARE — Controlling LLM-Generated Policies through Auditable Review of Evidence in Scientific Experimentation

·9 min read·Guanyu Liu, Weiyi Kong, Zeyu Wang et al.

This paper addresses a critical challenge in deploying large language models (LLMs) to control real-world scientific experiments, specifically high-throughput experimentation (HTE)

researchlarge-language-modelsscientific-experimentationhigh-throughput-experimentationauditable-control

Read note → Source paper ↗

From THESAN-ZOOM to JWST — Predicting ionizing photon escape and the rise of UV-bright reionization sources

research note

From THESAN-ZOOM to JWST — Predicting ionizing photon escape and the rise of UV-bright reionization sources

·9 min read·Zebedee Summerfield, William McClymont, Sandro Tacchella et al.

This paper addresses the critical astrophysical problem of understanding the sources and evolution of cosmic reionization, focusing on predicting the escape fraction (f_esc) and escape rate (\dot{N…

researchcosmological-simulationsepoch-of-reionizationmachine-learningrandom-forest

Read note → Source paper ↗

Articles are CC BY 4.0 — feel free to quote with attribution