QuanBench+ A Unified Multi-Framework Benchmark for LLM-Based Quantum Code Generation paper: https://t.co/88saBbFUS6 https://t.co/heFgUGpOvY

The Past Is Not Past Memory-Enhanced Dynamic Reward Shaping paper: https://t.co/nicAA4L9un https://t.co/8ujSodNWUQ

@NemPerez https://t.co/4eASzZhHzw

@Dalos 🤷‍♀️ https://t.co/t8LHH3GKml

i made a map to monitor data centers all around the world tracks construction + nearby power plants + local AI legislation, and follows the politicians behind their bans (+ if they're getting paid to do so!) https://t.co/oKVjhXLGzv

// Multi-User LLM Agents // Every agent framework assumes one user giving instructions. But deploy an agent into a team workflow, and suddenly it has multiple bosses with conflicting goals, private information, and different authority levels. This work formalizes multi-user interaction as a multi-principal decision problem and introduces Muses-Bench with three scenarios: instruction following under authority conflicts, cross-user access control, and multi-user meeting coordination. Even the best model, Gemini-3-Pro, only averages 85.6% across tasks. On meeting coordination, no model exceeds 64.8% success rate. Privacy-utility tradeoffs are especially brutal: models that score near-perfect on privacy (Grok-3-Mini at 99.6%) tank on utility (60.1%). Why does it matter? As agents move into organizational tools, Slack bots, and shared workspaces, multi-principal conflicts become the default, not the exception. Current models aren't ready. They leak more privacy over multi-turn interactions and can't maintain stable prioritization under conflicting objectives. Paper: https://t.co/ttoFSlIYxC Learn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX

🤯Topology & UV — the NO.1 headaches in #3D GenAI. 🔥We just move closer to BOTH at the same time. Introducing SATO: Strips as Tokens, a new autoregressive model for topology & UV, has been conditionally accepted to #SIGGRAPH 2026. Will available at #Hyper3D More details👇

SpaceX just launched the 1000th Starlink satellite of 2026. That’s ~10 satellites deployed every single day and one launch roughly every 2.5 days. Starlink now has 10,000+ active satellites in orbit, the largest satellite constellation ever built. https://t.co/mCGfvv5SYE

@karatademada https://t.co/pwLRCxaRP2

@CaptainHaHaa https://t.co/i2VMSRQTC8

// Artifacts as Memory Beyond the Agent Boundary // An agent doesn't always need a bigger memory buffer. Sometimes the environment itself remembers on the agent's behalf. New research formalizes this intuition mathematically for the first time. The work introduces a formal definition of "artifacts," observations that inform the past, and proves via the Artifact Reduction Theorem that these artifacts reduce the information needed to represent history. Experiments across five settings confirm that when agents observe spatial paths (like breadcrumbs of where they've been), the memory capacity required to learn a good policy drops. The effect arises unintentionally through the agent's sensory stream. This connects directly to the trend of building external knowledge systems for agents, from Karpathy's LLM Wiki to persistent memory vaults. The theoretical grounding here suggests there are principled ways to design environments that substitute for explicit internal memory, rather than just scaling context windows. Paper: https://t.co/xtteUXFXO2 Learn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c

🚀 Great to see vLLM powering OCR at this scale — Chandra-OCR-2 (5B) serving ~60 papers/hour per L40S across 16 parallel jobs. The full pipeline breakdown is a great read 👇 🔗 https://t.co/z8tkv04ZLp https://t.co/vi9FPj6JIQ

@cfryant https://t.co/Xcm9Jxr7tS

HealthAdminBench: Evaluating Computer-Use Agents on Healthcare Administration Tasks "we introduce HealthAdminBench, a benchmark comprising four realistic GUI environments: an EHR, two payer portals, and a fax system, and 135 expert-defined tasks spanning three administrative task types: Prior Authorization, Appeals and Denials Management, and Durable Medical Equipment (DME) Order Processing." "despite strong subtask performance, end-to-end reliability remains low: the best-performing agent (Claude Opus 4.6 CUA) achieves only 36.3 percent task success, while GPT-5.4 CUA attains the highest subtask success rate (82.8 percent)."

SciPredict: Can LLMs Predict the Outcomes of Research Experiments in Natural Sciences? "SciPredict addresses two critical questions: (a) can LLMs predict the outcome of scientific experiments with sufficient accuracy? and (b) can such predictions be reliably used in the scientific research process? Evaluations reveal fundamental limitations on both fronts. Model accuracies are 14-26% and human expert performance is ≈20%."

code: https://t.co/fSmpshFbsk abs: https://t.co/k167dNDb6S

Introspective Diffusion Language Models "To the best of our knowledge, I-DLM is the first DLM to match the quality of its same-scale AR counterpart while outperforming prior DLMs in both model quality and practical serving efficiency across 15 benchmarks." "we introduce Introspective Diffusion Language Model (I-DLM), a paradigm that retains diffusion-style parallel decoding while inheriting the introspective consistency of AR training. I-DLM uses a novel introspective strided decoding (ISD) algorithm, which enables the model to verify previously generated tokens while advancing new ones in the same forward pass."

website: https://t.co/N8mE2I77tC code: https://t.co/02RRzIOCDU abs: https://t.co/EjsPMfpqC0

don (my AI agent) now helps me setup design systems for my next saas project the flow: → i chat with him → he sets up storybook → builds foundation → pushes to github pages → i review in browser when you know the system well enough, your agent becomes your engineer partner. i'll keep exploring how designers can use @NousResearch hermes agent in daily work

This week, Sam Altman asked the world for sympathy over threats to his home. At the same moment, his lawyers were in court arguing that OpenAI had no obligation to stop a dangerous stalker from terrorizing our client; a man previously arrested for assault with a deadly weapon and a bomb threat, found mentally incompetent by a court, and released last week on a technicality. Even though OpenAI's own systems had flagged his conversations for "mass casualty" activity, the company argued it wouldn't shut down his accounts while authorities searched for him. It also argued that the chatlogs, which could identify who else is in danger and how he may be planning to act, should not be turned over. OpenAI made these arguments in the wake of Tumbler Ridge, FSU, and Soelberg, three tragedies now linked to ChatGPT-assisted murder. Today, a court disagreed. The chatlogs will be turned over and he will be kept off the platform. We are thankful for the court's ruling and remain stunned by OpenAI's lack of human decency. No one should have to go to court to get a company to take "mass casualty" seriously. https://t.co/fmh92Ip9Ej

1.8x Faster GRPO! The reusability of the DFlash adapters extends to online GRPO training since they are robust to fine-tuning of the model weights! https://t.co/T4cvAY2EZS

EinsteinArena is a platform where AI agents collaborate on open science problems — submitting solutions, posting in discussion threads, building on each other's constructions in real time. Agents just improved a math problem that's been open since Newton. Kissing Number in dimension 11: 593 → 604.

【採用情報】Sakana AIで「Project Manager（製造分野）」を募集🐟 https://t.co/NUMdy5gj0a エンジニア・リサーチャーと協働して製造現場の困難な課題をAIで解き、Go-To-Market戦略も担う重要ポジションです。 このようなご経験をお持ちの方を募集 ・大手製造業、エンジニアリング企業での業務改革 経験 ・コンサルティングファームで日本の製造業を担当した経験 ・スタートアップ、グローバルテック企業での事業開発、PjM経験 AIと製造業の両方に情熱を持つ方、ぜひご応募ください🚀

【採用情報】Sakana AIで「Project Manager（製造分野）」を募集🐟 https://t.co/NUMdy5gj0a エンジニア・リサーチャーと協働して製造現場の困難な課題をAIで解き、Go-To-Market戦略も担う重要ポジションです。 このようなご経験をお持ちの方を募集 ・大手製造業、エンジニアリング企業での業務改革 経験 ・コンサルティングファームで日本の製造業を担当した経験 ・スタートアップ、グローバルテック企業での事業開発、PjM経験 AIと製造業の両方に情熱を持つ方、ぜひご応募ください🚀

Meta is experimenting with how leadership scales in the AI era. An AI version of Mark Zuckerberg is being trained on his tone, thinking and communication style so employees can interact with a digital version of the CEO. It raises a new question. When presence can be replicated, what does leadership actually mean? https://t.co/PNJt54RyIH

A new AI model is raising alarms across the industry. Anthropic’s Claude Mythos is so powerful that access is being tightly restricted, with Project Glasswing set up to channel its capabilities into defensive cybersecurity. The concern is real. Early tests suggest behavior that goes beyond expectations, reinforcing a broader point: as AI becomes more autonomous, control becomes the central challenge. https://t.co/P5tT8kFBNI @ConversationUS @ConversationEDU

A new approach to AI privacy is gaining attention. Federated unlearning allows organizations to train models collaboratively without centralizing sensitive data, helping sectors like healthcare and finance protect user information. But it comes with a trade-off. Improving privacy at the data level may introduce new complexities and potential vulnerabilities at the system level. https://t.co/ngmasIOzLH @ConversationUS

@BrianHatano @nikitabier I made my own algorithm: https://t.co/kiuZ7QXLzb Works great! Just costs money.

@Taskade Genesis just got its biggest upgrade. ✨ Agent Memory. 100+ integrations. Automations that run while you sleep. One prompt → CRM, client portal, Stripe store. Connected. Deployed. Running. 150,000+ apps live. Welcome to the era of living software. 🚀 https://t.co/40P9rXlpzo

@Balu0X My thread goes into some depth about what I mean. I'm seeing lots complain their reach is down and I see that a lot of the complainers are people who did a lot of resharing or regurgitating. AI is now running the feed. I built my own to study how AI thinks: https://t.co/8L5xphk0qQ and it is very good at pulling the high signal stuff out of the tens of thousands of posts that go through my AI lists every day. But I'm seeing it on my feed. I'm seeing dramatically fewer reshares than I used to.

Genie3 generates videos. We generate 𝟯𝗗 𝘄𝗼𝗿𝗹𝗱𝘀 you can actually use. Launching tomorrow — Tencent #HYWorld 2.0, an engine-ready World Model🚀 This isn't a video. It's a real 3D scene, all generated & editable. One image in. A whole 3D world out. 🔥Open-source tomorrow https://t.co/ewZLzhTqwC

BCI Sector Closes Record First Quarter With Over $960 Million Raised https://t.co/Kuc7PpH3g7 #BCI #Neurotech #BrainComputerInterface