The Semantic Scholar Open Data Platform

cs.DL 2026-07 conditional novelty 8.0 of 10

At ICLR, equally scored borderline papers from outside top-25 institutions are accepted less often, a gap concentrated in preprint-identifiable submissions; outcome tests find no evidence of a higher bar.

cs.CL 2026-07 unverdicted novelty 7.0 of 10

NLP authors show migration from *ACL flagship tracks (–19.2pp) to Findings (+14.8pp) and ML venues (+8.6pp), with new authors increasing ML share from 5% to 21% and causal inference indicating a citation premium drive...

cs.DL 2026-06 unverdicted novelty 7.0 of 10

Science and software impact each other through complementary strata, but sparse paper–repo linkage makes reuse–citation coupling gap-sensitive and prevents strong decoupling claims.

cs.AI 2026-08 conditional novelty 6.0 of 10

An LLM-assisted assessment tool, a provenance ontology, and an open knowledge graph holding 140 research integrity assessments of 95 randomised clinical trial publications.

quant-ph 2026-07 conditional novelty 6.0 of 10

Manual review of 127 NISQ papers plus automated scan of ~5000 QC papers finds ~25% code availability and ~65% execution failure among those with code, with concrete recommendations.

cs.DL 2026-07 conditional novelty 6.0 of 10

An editor-native platform unifies research, writing, and publishing in one LaTeX environment with compile-verified agents and patent-to-paper impact signals.

cs.AI 2026-06 unverdicted novelty 6.0 of 10

Xcientist is a research harness that externalizes an AI scientist's literature grounding, idea evolution, experiments, and repairs into auditable artifacts, demonstrated on memory, traffic forecasting, and PDE-solving tasks.

cs.DL 2026-03 unverdicted novelty 6.0 of 10

Joint LLM ranking of all citations within a paper identifies impactful references more accurately than isolated classification, gaining +9.5% accuracy and +8.3% F1.

cs.CL 2025-06 conditional novelty 6.0 of 10

A continued-pretrained ModernBERT encoder for biomedical and clinical text claims SOTA on several clinical NLP tasks, with caveats about data overlap between pretraining and evaluation.

cs.CL 2025-06 reject novelty 6.0 of 10

Computer science-only teams now supply a growing majority of societally-oriented AI research on arXiv, even though interdisciplinary teams remain more likely to produce such work.

cs.CL 2025-06 conditional novelty 6.0 of 10

A new 8TB openly-licensed text corpus trains 7B LLMs that are competitive with Llama 1/2, showing that performant models need not depend on unlicensed web data.

cs.HC 2025-02 conditional novelty 6.0 of 10

Interviews with 11 computing survey authors show that keeping narrative surveys up to date is valued but unrewarded, and that updates fall into empirical, structural, and interpretive types.

cs.CL 2025-01 conditional novelty 6.0 of 10

In human evaluations by three professional editors, GPT-4V captions for scientific figures were preferred over author-written captions and over captions from challenge-winning models.

cs.CL 2024-12 conditional novelty 6.0 of 10

Analysis of tens of thousands of papers shows NLP/AI research mixes 'dialogue' and 'dialog' with no clear trend, author, or context explanation.

cs.DL 2024-12 reject novelty 6.0 of 10

Citing a retracted paper is associated with a growing citation deficit that is larger for indirect citations and in lower-impact journals, a pattern the authors call 'attention escape'.

cs.DL 2026-01 conditional novelty 5.0 of 10

A new citation-purpose taxonomy applied to NLP+CSS papers finds that most out-of-discipline citations are shallow and that automated classification of citation purpose is not yet reliable.

cs.CL 2025-08 conditional novelty 5.0 of 10

MedSEBA is a RAG-based medical question-answering system that provides cited key arguments, per-study stance labels, and temporal consensus visualization, evaluated by a 10-person user study.

cs.CL 2025-05 conditional novelty 5.0 of 10

ClimateEval unifies 25 climate-related NLP tasks into one benchmark and shows that open-source LLMs gain from few-shot examples but lag on misinformation and fine-grained entity recognition.

cs.IR 2024-12 reject novelty 5.0 of 10

LLM recommendations of important AI research favor recent, well-cited, team-authored papers, but do not measurably over-represent male, white, or developed-country scholars relative to a human-curated benchmark.

cs.DL 2026-07 conditional novelty 4.0 of 10

AtomGPT's hybrid LLM+Semantic Scholar checker flags 94 of 100 confirmed hallucinated NeurIPS 2025 citations, driven mainly by author mismatch rather than title similarity.

cs.AI 2025-05 conditional novelty 4.0 of 10

A call to add olfaction, with standardized data and benchmarks, to the list of core modalities that embodied AI systems should sense and reason about.

cs.CL 2025-02 conditional novelty 4.0 of 10

With few-shot prompting, Llama 3.1 classifies paper titles and abstracts into five ORKG top-level fields at 0.82 accuracy, about 0.08 above a BERT baseline.

cs.SI 2025-01 unverdicted novelty 4.0 of 10

A web demo maps about 28,000 biomedical researchers and 1,179 datasets into a searchable 2D space and uses GPT-4o to explain collaborator and dataset recommendations.

cs.DL 2025-08 unverdicted novelty 2.0 of 10

A community perspective on how AI can support scholarly knowledge extraction, organization, and communication, with a proposed classification and ethical considerations.