News

Researchers Bruno Franco and Edson Scalabrin introduced Compositional Graph Similarity (CGS), a graph-based metric for evaluating Transformer models on the ReCOGS benchmark, and found that the lowest-scoring categories a…

Koharu Saeki and Daisuke Bekki from Ochanomizu University introduced Express, a web-based interactive development environment for the Japanese automated inference system lightblue, which shifts grammar development from e…

Researchers Chit-Fung Lam and Elaine Uí Dhonnchadha presented the development of Cantonese and Irish treebanks within the Parallel Grammar (ParGram) Project at the Third Workshop on the Bridges and Gaps between Formal an…

Judith Sieker and Sina Zarrieß, in a position paper presented at the Third Workshop on the Bridges and Gaps between Formal and Computational Linguistics (BriGap-3) in Paris, France, in July 2026, argue that misalignments…

Researchers Nanako Miyagawa, Hinari Daido, and Daisuke Bekki introduced Neural Wani, an integration of a lightweight LSTM-based neural model into the automated theorem prover wani for Dependent Type Theory, achieving a 1…

Researchers Artem Saloev, Erin Pacquetet, and Nicolas Ballier found that the ABX experiment with Mimi's semantic token codebook fails to capture the mapping of semantic tokens to phone realisations, but by realigning Mim…

Materialize Inc. released its July 2026 product update, adding blue-green cutovers for Kafka sources, AWS Glue Schema Registry support, PostgreSQL physical replica support, and Iceberg sinks on Google Cloud (public previ…

PostHog launched Replay Vision, an AI layer for session recordings that detects user struggles and automatically drafts fixes as pull requests, with two internal cases solved in a week. The tool identified a dead button …

Researchers from the University of Alicante and other institutions presented a scalable detection method for command-line obfuscation using a custom-trained small transformer-based model, which outperformed previous appr…

Priscilla Adenuga's paper, presented at the Second International Conference on Natural Language Processing and Artificial Intelligence for Cyber Security (NLPAICS) in June 2026 in Alicante, Spain, argues that prompt inje…

Researchers from the University of Alicante and other institutions introduced the Neuro-Cyber Exploitation and Reconnaissance Taxonomy (NeuroCERT), a framework mapping cybersecurity side-channel attacks and active vector…

A study presented at the Second International Conference on Natural Language Processing and Artificial Intelligence for Cyber Security (NLPAICS 2026) found that GPT-5-mini outperformed Qwen2.5-Coder-7B in generating test…

Researchers Ivan Martinez-Murillo, Robiert Sepúlveda-Torres, and Juan Pablo Consuegra-Ayala found that transformer-based fraud detection models trained only on English achieve near-perfect in-language performance (F1 ≈ 0…

Researchers Ryan Hyland, Lewis Newsham, and Daniel Prince found that persona prompting can systematically shift the responses of a pre-trained large language model along Schwartz's ten basic human value dimensions, as me…

Researchers at the Second International Conference on Natural Language Processing and Artificial Intelligence for Cyber Security reported that fine-tuning small language models on synthetic cybersecurity data yields mixe…

Rui Sousa-Silva, in a paper presented at the Second International Conference on Natural Language Processing and Artificial Intelligence for Cyber Security (NLPAICS) in Alicante, Spain, in June 2026, found that forensic l…

Granola CEO Chris Pedregal said his AI notetaker company, valued at $1.5 billion, refuses to sell employee transcripts to bosses, calling such surveillance "not a future we want to build towards." Pedregal predicted that…

Asian stocks rose Friday, with South Korea's Kospi Index surging over 14% as chip majors SK Hynix and Samsung Electronics rallied, reviving confidence in the artificial intelligence trade. The yen held gains after Japane…

A Cloudflare study measuring agent-readiness capabilities found that well-known files for agent discovery (llms.txt, sitemap.xml, MCP server cards) were consulted zero times by every AI model tested, but a single sentenc…

A software engineer recounts how mandatory AI training at their company led to the widespread use of large language models to generate business intelligence reports without verification, resulting in hallucinated content…