alchemy-utils 0.1a1
Release: alchemy-utils 0.1a1 Performance boost for DuckDB exports and CSV imports, see here.
Full feed
Feed 27,975 posts
Every story across every category, newest first. Each card links to the original publisher; daily-brief posts open as editorial pages.
NeuPAT: Neuron-aware Plasticity Allocation Tuning for Language-Preserving MLLMs
Multimodal expansion of large language models (LLMs) enables new perceptual capabilities but often compromises the language intelligence acquired during pretraining. In this work, we investigate this…
AI4AI at Test-Time: Strong-to-Weak Capability Transfer via Harnesses
Recent work on distillation transfers the capabilities of large models to smaller ones often by updating the latter's parameters, through teacher forcing, on-policy distillation, and related…
China was surprisingly decentralized in the 90s The mountains are high, the emperor is far away etc. relevant anecdote from LKY: you can protest, but …
China was surprisingly decentralized in the 90sThe mountains are high, the emperor is far away etc.relevant anecdote from LKY: you can protest, but not against the…
From Synthesis to Removal: Physics-Grounded Reflection Simulation and Diffusion-Based Video Dereflection
Videos captured through glass often contain reflections that degrade visual quality and interfere with downstream vision tasks. Although single-image reflection removal has been extensively studied, video…
MBA: Multimodal Benchmark and Agents for Real-World Business Ideation
Agentic systems powered by large language models (LLMs) have opened new opportunities for business ideation. Yet existing approaches remain confined to a text-only paradigm, despite the…
I didn't notice that DeepSeek has updated their status page with fine-grained service statistics
I didn't notice that DeepSeek has updated their status page with fine-grained service statistics
AutoWorldModel-Bench: A State-Centric Benchmark for Automated World-Model Research
World modeling is an unsettled field: architectures, training objectives, and state representations interact in complex ways, and no single recipe dominates across environments. This makes it…
AVA-Encoder: Towards Agent-Native Video Representation Learning
Creative agents still lack an effective way to learn from high-quality human films, limiting their ability to produce cinematic-grade videos. A key challenge is the absence…
StateFlow: Building, Evolving, and Accessing 3D World States for Previsualization
Previsualization is an intermediate layer between ideas and production in film, games, architecture, and urban design. It lets creators iteratively refine scenes, actions, cameras, and spatial-temporal…
RT Mark Witzke: “Zhu noted that China was not privatizing state firms but simply turning them into modern corporations, but Bush winked and replied c…
RT Mark Witzke“Zhu noted that China was not privatizing state firms but simply turning them into modern corporations, but Bush winked and replied confidently, “We know…
If we take the average jump gen to gen (50%) from 2009 to 2025, on 2030 we should have a GPU that is 3x times faster than a 5090. Do you think that will happen? Or on the worst case..
We also have the other cases, where it would look like this: Scenario for 3x 5090 perf Uplift/Gen Approximate year Best 70% ~2029 Historical Average 50.3%…
Measuring Activation Control in LLMs
TL;DRInspired by the introspective awareness and CoT controllability papers, we made a benchmark to measure how well models can control their activations while completing a simple…
RT ITmedia AI＋: 「Sakana Fugu」が無料で試せるように チャットAIをアップデート、Sakana AI https://www.itmedia.co.jp/aiplus/article/2608/13/2000000524/
RT ITmedia AI＋「Sakana Fugu」が無料で試せるように チャットAIをアップデート、Sakana AIhttps://www.itmedia.co.jp/aiplus/article/2608/13/2000000524/
RT ITmedia AI＋: 「Sakana Fugu」が無料で試せるように チャットAIをアップデート、Sakana AI https://www.itmedia.co.jp/aiplus/article/2608/13/2000000524/
RT ITmedia AI＋「Sakana Fugu」が無料で試せるように チャットAIをアップデート、Sakana AIhttps://www.itmedia.co.jp/aiplus/article/2608/13/2000000524/
RT Mike Lee: I’m not sure who made this. But it rocks.
RT Mike LeeI’m not sure who made this.But it rocks.
RT hardmaru: We just pushed a big update to Sakana Chat. No login required and free to use: https://chat.sakana.ai/ It is now powered by Fugu and our …
RT hardmaruWe just pushed a big update to Sakana Chat.No login required and free to use: https://chat.sakana.ai/It is now powered by Fugu and our updated Namazu…
We just pushed a big update to Sakana Chat. No login required and free to use: https://chat.sakana.ai/ It is now powered by Fugu and our updated Namaz…
We just pushed a big update to Sakana Chat.No login required and free to use: https://chat.sakana.ai/It is now powered by Fugu and our updated Namazu Japanese…
That's already 25% of their import cut induced by the Hormuz closure a few more years and they can make the import reduction permanent with EVs alone….
That's already 25% of their import cut induced by the Hormuz closurea few more years and they can make the import reduction permanent with EVs alone.…
Agent Safety Should Be a Runtime Contract
The dominant paradigm treats AI safety as a property to be instilled during model training via RLHF, DPO, or Constitutional AI. We argue this is structurally…
Spark-to-Paper: End-to-End Research Paper Generation as a Composable Skill
Turning a research idea into a complete paper requires more than text generation: the system must retrieve literature, design and execute experiments, revise claims according to…
ToolHazard: Scaling Adversarial Environments for Security Evaluation and Alignment of LLM-based Agents
Large language model (LLM) agents integrated with external tools are vulnerable to indirect prompt injections embedded in environmental states. However, existing studies largely rely on manually…
realistically, DSH drops in another ≈7-9 hours I predict that this will recontextualize the last two releases. Enough fog of war, we'll see how it is…
realistically, DSH drops in another ≈7-9 hoursI predict that this will recontextualize the last two releases. Enough fog of war, we'll see how it is *supposed*…
[AINews] SpaceXAI Grok 4.6 and Grok @Bot
AI teammate category just had its most significant new entrant yet