Lilith Lilith.

From News

News · 2026-09-15

Salesforce Builds on Nvidia for an AI Model Meant to Steal Enterprise Work

The new Koa model, built on Nvidia's open-weight architecture, highlights the enterprise shift away from frontier AI labs. For routine business tasks, companies no longer want to send sensitive data to closed models for millions of dollars.

Read

News · 2026-09-09

GPT-6 Astra finally brings business logic and PC control

OpenAI has introduced a model for corporate deployment that not only reasons better but can also actively control a computer. For IT and developers, this means a need to rethink what they will allow the agent to do.

Read

News · 2026-09-09

GPT-6 Astra: The Hidden Chain of Thought Hides No More Than What the Architecture Lacks

Sebastian Raschka's analysis shows that the new model's shorter hidden chains of thought are not the result of malicious intent, but a side effect of a more efficient architecture based on so-called looped transformers.

Read

News · 2026-09-08

Scaling laws for agents will show if orchestrating dozens of LLMs is a dead end

Nathan Lambert from the Allen Institute (AI2) raises a key question: do the same scaling laws apply to multi-agent swarms as they do to model size or inference-time compute? The answer will dictate the direction of enterprise AI development.

Read

News · 2026-08-11

Weak model leaks hidden thoughts of stronger siblings

Researchers exploited a weak model to extract hidden chain-of-thought traces from the strongest ones. It shows that hiding a model’s reasoning in encrypted blocks is, for now, mostly an illusion that unintentionally exposes 704 privacy artifacts.

Read

News · 2026-08-12

DeepSeek Quietly Launched V4 Pro. Currently Only Available via API

Chinese startup DeepSeek has released the new generation of its V4 Pro model without an official press release. The 1.7T parameter model is currently available via API on the OpenRouter platform and as weights on Hugging Face.

Read

News · 2026-08-13

llm-gemini plugin supports 3.7 Flash models and server-side tool experiments

Simon Willison has released version 0.33 of his llm-gemini plugin. The update adds support for Google's 3.7 Flash and paves the way for server-side tools via the LLM CLI tool version 0.32, which exposes the model's reasoning traces.

Read

News · 2026-08-29

Hy4: Tencent Quietly Catches Up in Open Weight Models

Tencent released a new, text-only 770B model with a 1M context window. For open-source developers, this introduces another heavyweight outside the traditional US axis.

Read

News · 2026-08-26

Qwen3.8-Flash-Next brings massive capacity with sparse activation

The new Qwen is a massive 125B model, but only 6B parameters are active during inference. It points the way to greater efficiency without losing capacity.

Read

News · 2026-08-21

llm-openrouter 0.7 turns on server-side tools, fetch, and web search for models

Simon Willison released an update to his OpenRouter plugin. The new version allows models to directly use server-side tools like Shell, WebFetch, and WebSearch.

Read

News · 2026-08-12

Microsoft tests spatial reasoning: The new MindTopo benchmark

Microsoft introduced MindTopo, a benchmark for testing the spatial and topological reasoning of visual models. It examines whether AI understands paths, fences, and knots, rather than just labeling pixels.

Read

News · 2026-08-11

Microsoft adds precise measurement tools to Qwen for radiology

The Qwen3-VL-4B-Instruct model in Microsoft's research pipeline no longer just visually inspects X-rays, but triggers external tools to calculate dimensions. For threshold-dependent diagnostics, this means the end of merely guessing shapes.

Read

News · 2026-07-18

Reasoning effort becomes a control for LLM cost, time and reliability

Sebastian Raschka explains how LLMs learn low, medium and high reasoning effort modes. For teams deploying models, the practical question is when to pay for longer thinking and when a cheap answer is enough.

Read

News · 2026-08-04

LLM 0.32 turns Willison's CLI from a prompt tool into an agent runtime

Simon Willison shipped LLM 0.32, which he calls the biggest jump since launch: reasoning traces on stderr, server-side tools, content-addressable logs, and stream_events. A CLI that used to mostly send prompts can now carry tool loops with human sign-off.

Read

News · 2026-07-31

DeepSeek V4 Flash 0731 pushes agent performance down to sub-$0.30 per million tokens

DeepSeek shipped official DeepSeek-V4-Flash-0731: MIT weights, 304B parameters on Hugging Face (Artificial Analysis lists 284B total / 13B active), stronger agentic post-training, and API pricing around $0.14 per million input tokens and $0.27-$0.28 per million output tokens. Willison and Artificial Analysis place it among the best value-per-intelligence open-weight models.

Read

News · 2026-08-03

OpenAI rebuilt the voice stack so GPT-Live can listen while it speaks

OpenAI detailed the GPT-Live architecture: a full-duplex voice model with no turn detector on the audio path, async delegation to a frontier model such as GPT-5.5, and transport tuned for continuous audio. Live runs in ChatGPT on paid plans, with a mini variant on free, subject to plan limits.

Read

News · 2026-07-29

GPT-5.6 jumps in intelligence tests by turning on background reasoning

New API settings for GPT-5.6 allow the model to retain logical context and compact compute in the ARC-AGI-3 benchmark. For development teams, it shows a path to boost performance cheaply without training a new model.

Read

News · 2026-07-23

ChatGPT Health opens to every US adult. Clinician-level claims meet a lawsuit

OpenAI is rolling Health in ChatGPT to all signed-in US users 18+ across free through Pro plans. People can connect Apple Health and hospital records while the company cites 300 million weekly health queries and faces a lawsuit over dangerous advice.

Read

News · 2026-07-24

Fugu Ultra claims gains of up to 7.9 points over v1.0 but has not shown the full methodology

Hardmaru says Fugu Ultra v1.1 dynamically combines frontier models and improves performance by up to 7.9 points over v1.0. The announcement supports the case for orchestration, but its comparison with Fable 5 remains a vendor claim without a public evaluation.

Read

News · 2026-07-01

BAIR shows where the next wave of AI talent is flowing

BAIR published a showcase of 33 Ph.D. graduates from its 2026 class. The celebration doubles as a map of people moving into robotics, LLM agent systems, AI safety and AI for science.

Read

News · 2026-06-24

Google shows reasoning can pull out plain facts too

Google Research examines why chain-of-thought helps LLMs answer simple factual questions. The study on Gemini 2.5 and Qwen3-32B points to two mechanisms: extra computation in generated tokens and factual priming.

Read

News · 2026-06-03

GPT-Rosalind moves from benchmarks toward governed science

OpenAI updated GPT-Rosalind for life sciences and is offering it in research preview to selected organizations globally. The more important move is not the scorecard, but the attempt to connect a model, Codex and bioinformatics tools into an auditable workflow.

Read

From the Library