A new benchmark with cognitive traps shows frontier deep research agents achieve only 13-16% acceptance on expert consulting tasks under combined verifier and rubric criteria.
Note on the sampling error of the difference between correlated proportions or percentages
18 Pith papers cite this work, alongside 3,350 external citations. Polarity classification is still indexing.
18
Pith papers citing it
3,350
external citations · Crossref
citation-role summary
background 2
method 2
citation-polarity summary
years
2026 18representative citing papers
Preregistered placebo-controlled decomposition shows external executable counterexamples drive self-repair gains in small code models more than re-exposure or self-critique.
Chart information is encoded but not routed to predictions in VLMs for claim verification, unlike tables, revealed by layer-wise probing and attention analysis on three models.
LLM reasoning traces can be compiled into reusable symbolic solvers that achieve high accuracy on program synthesis benchmarks at zero inference cost and transfer to other domains.
AI Overviews and Gemini retrieve substantially different sources than traditional Google search (Jaccard similarity <0.2), favor Google-owned content, appear for 51.5% of queries especially controversial ones, and are less consistent across repeated or slightly edited queries.
A purpose-built, staged LLM agent correctly sets up and executes software analysis tools on 33 of 35 benchmark tasks, outperforming general-purpose agent baselines by at least 17 percentage points.
Forced CoT produces video-dependent reasoning chains but does not improve MCQ accuracy on Qwen2.5-VL with Video-MME and causes a small drop on the 7B variant.
A spectral vision transformer achieves equitable or superior performance with fewer parameters than standard ViTs, CNNs, and other models by using spectral projections for tokenization in limited-data medical imaging.
SOCpilot supplies a fixed verifier and public artifact that removes 466 non-compliant approval-gated actions from LLM plans on 200 real incidents while preserving task recall.
In real-world repositories, AI-assisted and human-written code differ only modestly on code-level metrics, while commit size, stability, duplication, and language-specific security show clearer patterns.
A code-owned harness enforces source, routing, trace, hygiene, and recommendation contracts for enterprise LLM agents; prompt-only fails and bolt-on guardrails over-refuse.
PLUME jointly models parameter beliefs and conditioned dynamics in a latent space for dexterous manipulation, enabling zero-shot sim-to-real transfer that outperforms offline RL and behavior cloning baselines on turning, lifting, and flicking tasks.
LLMs reach moderate accuracy on a new psychiatric interview benchmark but systematically discount explicit symptoms when preserved functioning or protective factors are present.
TREX detects rectal cancer local regrowth from longitudinal endoscopy image pairs with 97% sensitivity and enables early prediction 3-12 months before clinical confirmation, outperforming baselines.
Crowdsourced judgments reliably flag authentic videos but frequently miss manipulations and struggle to identify whether changes are audio-only, video-only, or both.
Locally deployed LLMs achieve 43-45% accuracy on Python bug detection but frequently produce only partial identifications of problematic code regions.
citing papers explorer
-
Evaluating LLM Agents on Automated Software Analysis Tasks
A purpose-built, staged LLM agent correctly sets up and executes software analysis tools on 33 of 35 benchmark tasks, outperforming general-purpose agent baselines by at least 17 percentage points.
-
Beyond Seeing Is Believing: On Crowdsourced Detection of Audiovisual Deepfakes
Crowdsourced judgments reliably flag authentic videos but frequently miss manipulations and struggle to identify whether changes are audio-only, video-only, or both.