# forck.live — 22–28 January 2024 > AI lab announcements, news reports and podcasts. forck.live follows what companies announce about their own products. Press may corroborate an official story; it never leads and never appears as an item of its own. A missing date stays missing. A title is the publisher's words — where a company does not publish in English, an English rendering is listed separately as `Title in English` and named as this site's translation, never as theirs. This file is one issue of the weekly archive: every confirmed first-party announcement of ISO week 2024-W04, 22–28 January 2024, newest first. The window never moves, so this URL always names the same seven days — but a post published inside them and detected later is added when it arrives, so a closed week can still gain entries. Every entry is one first-party announcement, and carries the same eight fields in this order before any others: Company · Announced · Category · Coverage · Announcement · Group · Source · Record `Source` is the company's own page the announcement was traced to and `Record` is this site's page for it. Both are safe to cite. `Coverage` is a number of outlets, or `not counted` — which means this record reached no count of its own, never that nobody wrote about it. `Category` is this tracker's own analysis, and is authoritative wherever it is stated. `not stated` means either that nothing has analysed the post yet or that the analysis placed it outside this taxonomy; neither is a verdict about the announcement. `Announcement` is whether the post is news of something the company did, rather than a how-to, a customer story, a changelog listing or a staffing note. `Group` is which of this record's groups the entry falls in, carried on the entry itself so it holds where the groups are headings and where they are not. One of: `models`, `covered`, `announcements`, `routine`. Anything after `Record` is held detail that only some announcements carry. - This week's page: https://forck.live/week/2024-W04 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### Supabase — Create a Figma Clone app with Flutter and Supabase Realtime - Company: Supabase (supabase.com) - Announced: 2024-01-26 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://supabase.com/blog/flutter-figma-clone - Record: https://forck.live/items/21451-create-a-figma-clone-app-with-flutter-and-supabase-realtime - Subject: Supabase Supabase published a guide demonstrating how to combine its Realtime Broadcast feature with Flutter's CustomPainter to build a collaborative design board app. The tutorial covers drawing and moving shapes, syncing cursor positions and objects in real time, and persisting canvas state in a Postgres database. ### Hugging Face — An Introduction to AI Secure LLM Safety Leaderboard - Company: Hugging Face (huggingface.co) - Announced: 2024-01-26 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/leaderboard-decodingtrust - Record: https://forck.live/items/1952-an-introduction-to-ai-secure-llm-safety-leaderboard - Subject: Platform Hugging Face announces the release of the LLM Safety Leaderboard, powered by the HF leaderboard template, which focuses on safety evaluation for LLMs using the DecodingTrust framework. ### Alibaba — Introducing Qwen-VL - Company: Alibaba (alibaba.com) - Announced: 2024-01-25T05:33:00+00:00 - Category: model-update - Coverage: not counted - Announcement: yes - Group: models - Source: https://qwenlm.github.io/blog/qwen-vl/ - Record: https://forck.live/items/2320-introducing-qwen-vl - Subject: Qwen Alibaba announces the upgrade of the Qwen-VL series with two new enhanced versions, Qwen-VL-Plus and Qwen-VL-Max, featuring improved image reasoning, detail extraction, and support for high-resolution images. ### Upstage — Introducing Solar Mini: Compact yet Powerful - Company: Upstage (upstage.ai) - Announced: 2024-01-25 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://www.upstage.ai/blog/en/introducing-solar-mini-compact-yet-powerful - Record: https://forck.live/items/18487-introducing-solar-mini-compact-yet-powerful - Subject: Solar Pro / Solar Mini Upstage introduced Solar Mini, a pre-trained LLM that reached the top of the Hugging Face Open LLM Leaderboard in December 2023. The model uses fewer parameters than GPT-3.5 but is 2.5 times faster, and is built using a depth up-scaling method on a 32-layer Llama 2 architecture initialized with Mistral 7B weights. Solar Mini is publicly available under the Apache 2.0 license. ### Hugging Face — Hugging Face and Google partner for open AI collaboration - Company: Hugging Face (huggingface.co) - Announced: 2024-01-25 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/gcp-partnership - Record: https://forck.live/items/1953-hugging-face-and-google-partner-for-open-ai-collaboration - Subject: Platform Hugging Face and Google Cloud announced a strategic partnership to democratize AI, focusing on open science, open source, cloud, and hardware collaboration. ### Supabase — How pg_graphql works - Company: Supabase (supabase.com) - Announced: 2024-01-24 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://supabase.com/blog/how-pg-graphql-works - Record: https://forck.live/items/21452-how-pg-graphql-works - Subject: Supabase Supabase published a technical article explaining the internal workings of pg_graphql, the Postgres extension that powers its GraphQL API. The post covers how a GraphQL query is parsed and processed, from tokenization through AST generation, using the graphql-parser crate in Rust. ### LG AI Research — [NeurIPS 2023] Tabular Learning Research Trends Observed at NeurIPS 2023 - Company: LG AI Research (lgresearch.ai) - Announced: 2024-01-24 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.lgresearch.ai/blog/view?seq=392 - Record: https://forck.live/items/4874-neurips-2023-tabular-learning-research-trends-observed-at-neurips-2023 - Subject: EXAONE Research trends in tabular learning presented at NeurIPS 2023, including their own paper 'Binning as a Pretext Task' and overviews of other studies comparing neural networks and gradient-boosted decision trees. ### Hugging Face — Open-source LLMs as LangChain Agents - Company: Hugging Face (huggingface.co) - Announced: 2024-01-24 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/open-source-llms-as-agents - Record: https://forck.live/items/1954-open-source-llms-as-langchain-agents - Subject: Platform To build ReAct agents using open-source LLMs with LangChain, benchmarks models showing Mixtral outperforms GPT-3.5, and announces the release of the smolagents library. ### Alibaba — Introducing Qwen - Company: Alibaba (alibaba.com) - Announced: 2024-01-23T14:13:29+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://qwenlm.github.io/blog/qwen/ - Record: https://forck.live/items/2321-introducing-qwen - Subject: Qwen Alibaba provides an introduction to the Qwen series of large language models, four months after the initial release of Qwen-7B, outlining their work and objectives. ### LG AI Research — [NeurIPS 2023] Latest LLM Research Trends - Company: LG AI Research (lgresearch.ai) - Announced: 2024-01-23 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.lgresearch.ai/blog/view?seq=390 - Record: https://forck.live/items/4876-neurips-2023-latest-llm-research-trends - Subject: EXAONE LLM research trends at NeurIPS 2023, covering various papers on LLM performance, safety, tools, efficient learning, datasets, and reinforcement learning. No new product or model announcements are made. ### LG AI Research — [NeurIPS 2023] RLHF and RLAIF Research Trends and Noteworthy Papers - Company: LG AI Research (lgresearch.ai) - Announced: 2024-01-23 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.lgresearch.ai/blog/view?seq=391 - Record: https://forck.live/items/4875-neurips-2023-rlhf-and-rlaif-research-trends-and-noteworthy-papers - Subject: EXAONE This blog post from LG AI Research summarizes research trends in RLHF and RLAIF presented at NeurIPS 2023, highlighting two noteworthy papers on Direct Preference Optimization and Fine-Grained Human Feedback. ### Supabase — Getting started with Laravel and Postgres - Company: Supabase (supabase.com) - Announced: 2024-01-22 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://supabase.com/blog/laravel-postgres - Record: https://forck.live/items/21453-getting-started-with-laravel-and-postgres - Subject: Supabase Supabase published a guide on setting up a Laravel application with a Supabase Postgres database, covering project creation, authentication via Laravel Breeze, and database connection configuration. ### LG AI Research — [NeurIPS 2023] Introduction to Key Research Topics and Notable Papers - Company: LG AI Research (lgresearch.ai) - Announced: 2024-01-22 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.lgresearch.ai/blog/view?seq=382 - Record: https://forck.live/items/4877-neurips-2023-introduction-to-key-research-topics-and-notable-papers - Subject: EXAONE LG AI Research blog post summarizing key research topics and notable papers presented at NeurIPS 2023, covering Computer Vision, Large Language Models, Large Multimodal Models, Diffusion Models, and Reinforcement Learning, with highlights of specific papers such as Hummingbird, SiamMAE, and Toolformer.