@Techmeme@techhub.social2024-05-01 10:36:06
DOJ v. Google: Microsoft invested in OpenAI over fears of falling behind Google; Kevin Scott said he was "very, very worried" in a 2019 email to Satya Nadella (Bloomberg)
https://www.bloomberg.com/news/articles/20
@arXiv_csCL_bot@mastoxiv.page2024-05-01 06:48:59
Do Large Language Models Understand Conversational Implicature -- A case study with a chinese sitcom
Shisen Yue, Siyuan Song, Xinyuan Cheng, Hai Hu
https://arxiv.org/abs/2404.19509 https://arxiv.org/pdf/2404.19509
arXiv:2404.19509v1 Announce Type: new
Abstract: Understanding the non-literal meaning of an utterance is critical for large language models (LLMs) to become human-like social communicators. In this work, we introduce SwordsmanImp, the first Chinese multi-turn-dialogue-based dataset aimed at conversational implicature, sourced from dialogues in the Chinese sitcom $\textit{My Own Swordsman}$. It includes 200 carefully handcrafted questions, all annotated on which Gricean maxims have been violated. We test eight close-source and open-source LLMs under two tasks: a multiple-choice question task and an implicature explanation task. Our results show that GPT-4 attains human-level accuracy (94%) on multiple-choice questions. CausalLM demonstrates a 78.5% accuracy following GPT-4. Other models, including GPT-3.5 and several open-source models, demonstrate a lower accuracy ranging from 20% to 60% on multiple-choice questions. Human raters were asked to rate the explanation of the implicatures generated by LLMs on their reasonability, logic and fluency. While all models generate largely fluent and self-consistent text, their explanations score low on reasonability except for GPT-4, suggesting that most LLMs cannot produce satisfactory explanations of the implicatures in the conversation. Moreover, we find LLMs' performance does not vary significantly by Gricean maxims, suggesting that LLMs do not seem to process implicatures derived from different maxims differently. Our data and code are available at https://github.com/sjtu-compling/llm-pragmatics.
@MrBerard@pilote.me2024-02-29 10:07:29
Mastodon is predicated on community instances, and built for following/engagement with hashtags over accounts.
But if we've learnt anything from Late Capitalism, is that just because something is not needed it doesn't mean you can't build your whole business model around it.
🤦♂️
Newsmast brings curated 'communities' to the open source Twitter/X alternative Mastodon | TechCrunch
@arXiv_csHC_bot@mastoxiv.page2024-05-01 07:17:29
A Framework for Leveraging Human Computation Gaming to Enhance Knowledge Graphs for Accuracy Critical Generative AI Applications
Steph Buongiorno, Corey Clark
https://arxiv.org/abs/2404.19729 https://arxiv.org/pdf/2404.19729
arXiv:2404.19729v1 Announce Type: new
Abstract: External knowledge graphs (KGs) can be used to augment large language models (LLMs), while simultaneously providing an explainable knowledge base of facts that can be inspected by a human. This approach may be particularly valuable in domains where explainability is critical, like human trafficking data analysis. However, creating KGs can pose challenges. KGs parsed from documents may comprise explicit connections (those directly stated by a document) but miss implicit connections (those obvious to a human although not directly stated). To address these challenges, this preliminary research introduces the GAME-KG framework, standing for "Gaming for Augmenting Metadata and Enhancing Knowledge Graphs." GAME-KG is a federated approach to modifying explicit as well as implicit connections in KGs by using crowdsourced feedback collected through video games. GAME-KG is shown through two demonstrations: a Unity test scenario from Dark Shadows, a video game that collects feedback on KGs parsed from US Department of Justice (DOJ) Press Releases on human trafficking, and a following experiment where OpenAI's GPT-4 is prompted to answer questions based on a modified and unmodified KG. Initial results suggest that GAME-KG can be an effective framework for enhancing KGs, while simultaneously providing an explainable set of structured facts verified by humans.
@cowboys@darktundra.xyz2024-03-30 23:54:34
Latest Cowboys news relating to Ezekiel Elliott in free agency has bittersweet taste https://www.yardbarker.com/nfl/articles/latest_cowboys_news_relating_to_ezekiel_elliott_in_free_agency_has_…
@kennysmith@mstdn.social2024-02-23 04:52:28
“‘Being told that it can do all of these things is one thing, but actually seeing the capabilities, it was mind-blowing,’ he said in an interview with The Hollywood Reporter on Thursday, noting that his productions might not have to travel to locations or build sets with the assistance of the technology.
“As a business owner, Perry sees the opportunity in these developments, but as an employer, fellow actor and filmmaker, he also wants to raise in the alarm.”
@Mediagazer@mstdn.social2024-02-23 02:10:34
An interview with studio owner Tyler Perry, who says he's putting an $800M studio expansion in Atlanta on hold after seeing the capabilities of OpenAI's Sora (Katie Kilkenny/The Hollywood Reporter)
https://www.hollywoodreporter.com/business
@arXiv_csRO_bot@mastoxiv.page2024-02-27 08:25:18
This https://arxiv.org/abs/2310.07968 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csRO_…
@leodurruti@puntarella.party2024-03-13 15:48:47
“Twelve of Israel’s most prominent human rights organisations have signed an open letter accusing the country of failing to comply with the international court of justice’s (ICJ) provisional ruling that it should facilitate access of humanitarian aid into Gaza.”
https://www.
@arXiv_csRO_bot@mastoxiv.page2024-03-20 08:31:45
This https://arxiv.org/abs/2310.07968 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csRO_…
