https://megadose.ai/i/7995df22-9fee-4d79-b094-1c4fa7b8e471 Megadose en 2026-07-08T08:15:00+00:00 Chip startup SambaNova raised a $1B Series F led by General Atlantic at an $11B valuation and signs JPMorgan as a customer to deploy its chips for in-house AI (Bloomberg) https://megadose.ai/i/e151765c-aa65-436b-9b2c-b759f831cb0f Megadose en 2026-07-08T08:00:00+00:00 Hot French startup ZML releases free product to speed inference across lots of AI chips https://megadose.ai/i/f826a9dd-ca70-420f-9030-7c1c150e0756 Megadose en 2026-07-08T07:16:00+00:00 AI chip maker SambaNova raises $1B at $11B valuation, 5 months after last mega round https://megadose.ai/i/06395621-e99b-47ca-9454-a894bd456078 Megadose en 2026-07-08T05:05:51+00:00 Paris-based Skello, which provides AI-powered HR tools for frontline workforce management, raised €200M led by Bridgepoint to support its European expansion (David Cendon Garcia/EU-Startups) https://megadose.ai/i/49deb231-b2b8-423b-a60c-530ace3bb461 Megadose en 2026-07-08T04:30:01+00:00 Nvidia and AI chip startup d-Matrix are combining their hardware in a new system to power AI models, the latest example of Nvidia partnering with rivals (Phoebe Liu/The Information) https://megadose.ai/i/cb4bdcd6-1b4e-4ff5-9222-8a2f598b5c27 Megadose en 2026-07-08T00:00:29+00:00 Vibe coding has escaped the terminal https://megadose.ai/i/0fce6fbb-88ad-4621-b5b9-d4956ae63377 Megadose en 2026-07-07T22:30:00+00:00 As part of Meta's Muse Image rollout, Instagram users with public accounts need to opt out to block AI generations of their content by other users (Reece Rogers/Wired) https://megadose.ai/i/c0647470-bad0-4dae-b741-a49663142bb6 Megadose en 2026-07-07T22:18:10+00:00 Meta rolls out Muse, a new AI image generator https://megadose.ai/i/1d4b9e26-0751-4e37-8fa5-295ec0b100d3 Megadose en 2026-07-07T22:00:01+00:00 Anthropic plans to lease a 16-story building in Lower Manhattan and double its NYC workforce to 1,000 people this year, as AI companies expand in the city (New York Times) https://megadose.ai/i/d9a522fd-4b73-442a-9e0b-cae825c0e8a1 Megadose en 2026-07-07T21:40:00+00:00 OpenAI Chief Futurist Joshua Achiam is leaving the company later this month after nearly nine years (Maxwell Zeff/Wired) https://megadose.ai/i/ae67e862-c9d0-4e46-9ffe-ecf0afffd7dd Megadose en 2026-07-07T20:30:04+00:00 Meta launches Muse Image in Meta AI, Instagram, and WhatsApp, and previews Muse Video, the first media generation models from its Superintelligence Labs (Meta) https://megadose.ai/i/d523cc49-850a-4079-bf87-2d03a55a077f Megadose en 2026-07-07T19:58:20+00:00 Microsoft joins AI cost-cutting trend by relying more on its own models https://megadose.ai/i/35931f68-3b02-4cab-a5c2-e6ccd27577f1 Megadose en 2026-07-07T18:17:30+00:00 Meta rolls out Muse Image, the first image generation model from its Superintelligence Labs, in Meta AI; it will also power new tools in Instagram and WhatsApp (Meta Newsroom) https://megadose.ai/i/3365f3e4-03cc-4678-8bcf-55c29bd9c630 Megadose en 2026-07-07T17:59:50+00:00 ELSA3D: Elastic Semantic Anchoring for Unified 3D Understanding and Generation https://megadose.ai/i/28195f5e-7a14-47b5-97a3-6887aa5df0b4 Megadose en 2026-07-07T17:52:17+00:00 Graph Convolutional Attention: A Spectral Perspective on Graph Denoising and Diffusion https://megadose.ai/i/b36dea35-a854-45f9-b293-422030ad9522 Megadose en 2026-07-07T17:47:26+00:00 On the feasibility of dependency parsing of non-human sequences without a gold standard. Is evaluation possible in other species? https://megadose.ai/i/12bd7743-20e1-4755-92b3-e4d7f32fb0b5 Megadose en 2026-07-07T17:43:36+00:00 Hierarchical Acoustic-Semantic Modeling: Modality Separation and Semantic Coherence for Full-Duplex SLMs https://megadose.ai/i/849f1465-7494-4b7d-ad76-aec662907d34 Megadose en 2026-07-07T17:39:06+00:00 GraphBU: MILP Instance Generation with Graph-Native Block Units https://megadose.ai/i/acaa8863-de4d-4532-81f2-bc2aa58f7e57 Megadose en 2026-07-07T17:38:47+00:00 The Large Cancer Assistant (LCA): A Model-Agnostic Orchestration Framework for Scalable Clinical Decision Support in Oncology https://megadose.ai/i/ac2adb02-a138-40d9-9a06-e171cdbf2f54 Megadose en 2026-07-07T17:32:36+00:00 RSF-GLLM: Bridging the Semantic Gap in Multi-Hop Knowledge Graph QA via Recurrent Soft-Flow and Decoupled LLM Generation https://megadose.ai/i/5fdf4e03-d16f-43b4-a272-d3cd35f6d1ba Megadose en 2026-07-07T17:29:01+00:00 DepthWeave-KV: Token-Adaptive Cross-Layer Residual Factorization for Long-Context KV Cache Compression https://megadose.ai/i/13cc0d84-7223-4cba-997c-3991845deb0c Megadose en 2026-07-07T17:27:59+00:00 Bridging Physical Reasoning and Task Generalization via Visual Action Outcome Reasoning Alignment https://megadose.ai/i/84d77f1e-94c5-47b4-b844-343525cdeb5e Megadose en 2026-07-07T17:26:28+00:00 FreqDepthKV: Frequency-Guided Depth Sharing for Robust KV Cache Compression in Long-Context LLM Inference https://megadose.ai/i/bc8ea929-8a38-465b-ae8b-8e4fa7c94488 Megadose en 2026-07-07T17:18:44+00:00 FootsiesGym: A Fighting Game Benchmark for Two-Player Zero-Sum Imperfect-Information Games https://megadose.ai/i/296f2ea9-355a-4dc7-ab66-bb8158d5ef5f Megadose en 2026-07-07T17:09:36+00:00 DynaKRAG: A Unified Framework for Learnable Evidence Control in Multi-Hop Retrieval-Augmented Generation https://megadose.ai/i/406c8c15-bdfa-4815-8468-2064017a5f57 Megadose en 2026-07-07T17:04:44+00:00 RMISC: A Large-scale Real-world Multivariate Corpus for Time Series Foundation Models https://megadose.ai/i/166d446d-5d44-4e27-8477-095e1f14edaf Megadose en 2026-07-07T17:03:55+00:00 Doomed from the Start: Early Abort of LLM Agent Episodes via a Recall-Controlled Probe Cascade https://megadose.ai/i/d060251d-8bce-442c-aec1-1c7b5ff4850b Megadose en 2026-07-07T16:55:39+00:00 Pitwall: Faithful Natural-Language Race-Strategy Briefings from a Calibrated Real-Time Monte Carlo Engine https://megadose.ai/i/ce60d747-8531-4ed1-a17d-b38eff64c508 Megadose en 2026-07-07T16:46:46+00:00 AirflowAttack: Thermal-Airflow Adversarial Perturbations against Infrared Remote-Sensing Vision-Language Models https://megadose.ai/i/e38558dd-187b-4605-920a-1bddff0aecd0 Megadose en 2026-07-07T16:43:05+00:00 Data Analysis in the Wild: Benchmarking Large Language Models Against Real-World Data Complexities https://megadose.ai/i/d1416f73-ec5b-4a3a-9669-fc632c3b7d27 Megadose en 2026-07-07T16:41:25+00:00 Prompt-Adapter Context Routing for Parameter-Efficient Multi-Shot Long Video Extrapolation https://megadose.ai/i/c2b49069-ad52-4abd-ba42-96ab78df978d Megadose en 2026-07-07T16:30:09+00:00 Provable learning separation for predicting time-evolution of quantum many-body systems https://megadose.ai/i/a1fc999e-bb11-402b-a0df-45f758b1beb1 Megadose en 2026-07-07T16:30:02+00:00 Sources: Microsoft, looking to reduce AI costs, is starting to replace models from OpenAI and Anthropic with its MAI models in products like Excel and Outlook (Brody Ford/Bloomberg) https://megadose.ai/i/5e1e93e6-a7cd-46e1-835a-c9cdc611e672 Megadose en 2026-07-07T16:27:18+00:00 Claude Cowork expands to mobile and web https://megadose.ai/i/59ebf3ff-5886-49e0-9d4d-c58b1a0f579e Megadose en 2026-07-07T16:22:59+00:00 WordVoice: Explicit and Decoupled Multi-Dimensional Word-Level Control for LLM-Based TTS https://megadose.ai/i/4579fdfa-b7a9-47b9-964a-2049c8b22656 Megadose en 2026-07-07T16:11:30+00:00 Danus: Orchestrating Mathematical Reasoning Agents with Fact-Graph Memory https://megadose.ai/i/74f9b927-2809-4046-82b3-36a4bc058ea4 Megadose en 2026-07-07T16:10:44+00:00 Show HN: Rowboat – Open-source, local-first alternative to Claude Desktop https://megadose.ai/i/83545349-a7d2-4b1c-b281-46ccacc60677 Megadose en 2026-07-07T16:10:02+00:00 Anthropic expands Claude Cowork to web and mobile platforms in beta for Max plan subscribers, and says 90%+ of Cowork usage is unrelated to software development (David Gewirtz/ZDNET) https://megadose.ai/i/50d3c920-2a2c-4938-af4a-d0bcd364685c Megadose en 2026-07-07T16:05:20+00:00 Tangos, which uses AI to conduct financial crime investigations at scale, raised a $20M seed led by Red Dot (Mike Wheatley/SiliconANGLE) https://megadose.ai/i/8315fb48-653f-4ebe-92ce-40b4430111a2 Megadose en 2026-07-07T15:59:59+00:00 TILDE: TILt-based Distributional Erasure for Concept Unlearning https://megadose.ai/i/dcc58ae7-9af5-4595-b462-4eca092ec350 Megadose en 2026-07-07T15:43:22+00:00 An Experimental Design Approach to Evaluating Agentic AI's Autonomous Model Discovery https://megadose.ai/i/acdbf6fe-b497-4e8a-8e79-2414af7559a0 Megadose en 2026-07-07T15:41:22+00:00 RuBench: A Repository-Level Agentic Coding Benchmark with Natively Authored Russian Task Specifications https://megadose.ai/i/c2a9cccf-2dbd-464d-b0a0-bbc1a7afafc6 Megadose en 2026-07-07T15:21:20+00:00 A Function-Space Dichotomy for Compositional Learning: Exponential Sub-Optimality of the Neural Tangent Kernel https://megadose.ai/i/b92a452c-fe87-417b-bf41-df3e12098b6d Megadose en 2026-07-07T15:10:15+00:00 Training-Free Acceleration for Vision-Language-Action Models with Action Caching and Refinement https://megadose.ai/i/fd8a47c1-60e2-4bfa-bd38-76d6e935fe8e Megadose en 2026-07-07T14:47:46+00:00 TopoBrick: Agentic Topology Sampling of Exogenous Variables for Zero-Shot Building IoT Forecasting https://megadose.ai/i/1e76c802-32bb-4d4a-9e66-83e27b35a48b Megadose en 2026-07-07T14:45:42+00:00 Physics-Informed Neural Embeddings of PDE Solution Families https://megadose.ai/i/7ba0ff81-8df2-4a08-a9f9-e73f4a7499e6 Megadose en 2026-07-07T14:39:59+00:00 Harnessing Code Agents for Automatic Software Verification https://megadose.ai/i/49c7a31c-8734-44d9-93a0-b3584b385960 Megadose en 2026-07-07T14:25:26+00:00 Driving the Wrong Way: Leveraging Interpretability in End2End Autonomous Driving Models https://megadose.ai/i/3167aac5-3708-472e-ba68-8ee48e498f24 Megadose en 2026-07-07T14:25:09+00:00 Estimating Uncertainty from Reasoning: A Large-Scale Study of Multi- and Crosslingual MCQA Performance in LLMs https://megadose.ai/i/935ec622-3ed1-4781-82f0-a562e043c4a0 Megadose en 2026-07-07T14:25:05+00:00 DT-Guard: Intent-Driven Reasoning-Active Training for Reasoning-Free LLM Safety Guardrail https://megadose.ai/i/c2819f10-535b-4fef-bd46-2fc59393182f Megadose en 2026-07-07T14:20:00+00:00 Dithered Gaussian Mechanism for Randomness-Efficient Differential Privacy https://megadose.ai/i/611c6d7e-1ac8-4f1f-a55d-78e68cb97bb4 Megadose en 2026-07-07T14:11:54+00:00 Token-Based Dual-view Fusion and Adaptation of Large Vision Models for Breast Cancer Classification https://megadose.ai/i/93281cf9-7579-4064-9325-d10a5e740f17 Megadose en 2026-07-07T13:35:15+00:00 Solos unveils the AirGo A6 camera-less smart glasses, cutting the weight to 19g from 36-40g of the AirGo A5, and V2 privacy accessories like a clip-on shield (Andrew Liszewski/The Verge) https://megadose.ai/i/7733803a-b743-40a8-b0ab-a7ee5970abf2 Megadose en 2026-07-07T12:55:02+00:00 Norm, which runs an AI law firm that provides AI-powered legal services to clients alongside human lawyers, raised $120M led by Khosla at a $1.2B valuation (Guinevere Grant/Bloomberg) https://megadose.ai/i/b2a78c7e-2676-4947-bef4-794a624142d2 Megadose en 2026-07-07T12:25:01+00:00 Sources: Amazon is looking to raise at least $25B from a US dollar bond sale to fund its AI infrastructure investments; the size could increase based on demand (Bloomberg) https://megadose.ai/i/f5f38276-261b-465d-9163-a59f6030860b Megadose en 2026-07-07T12:16:00+00:00 SpaceXAI launches 21 new Grok Voices on xAI Console https://megadose.ai/i/fa45d703-54cd-47bc-90a9-5979e3dfb7d4 Megadose en 2026-07-07T12:00:00+00:00 Savi’s app aims to protect consumers from realistic AI scams like kidnappers demanding ransom https://megadose.ai/i/59b773b9-85aa-484c-9d56-a0f69d733065 Megadose en 2026-07-07T11:25:01+00:00 Paris-based UMA, founded by ex-Tesla Optimus scientist Rémi Cadene and ex-Google DeepMind researcher Pierre Sermanet, demos its Northstar AI humanoid robot (Benoit Berthelot/Bloomberg) https://megadose.ai/i/b0ac8db8-459a-4ee3-8ea3-f967ff942905 Megadose en 2026-07-07T10:45:00+00:00 Sources: DeepSeek is in the early stages of developing its own AI chip designed for inference, in a bid to reduce its reliance on Nvidia and Huawei hardware (Reuters) https://megadose.ai/i/2c206a97-bffb-470a-81d7-ccd969aa2faf Megadose en 2026-07-07T10:00:01+00:00 UK-based AI infrastructure startup Nscale secures a $900M line of credit to expand its data center buildout across Europe, the US, and the Asia Pacific (Mauro Orru/Wall Street Journal) https://megadose.ai/i/7036b3df-2868-4229-b0ec-5d62107667fc Megadose en 2026-07-07T00:00:00+00:00 SIEVE: Structure-Aware Data Selection for Imitation Learning with VLA Models https://megadose.ai/i/147bb08a-2b11-4ad7-bda7-cffbf15c1123 Megadose en 2026-07-07T00:00:00+00:00 AlayaWorld: Long-Horizon and Playable Video World Generation https://megadose.ai/i/a796ebb0-774c-4ecb-99b9-9ecfe9229411 Megadose en 2026-07-07T00:00:00+00:00 Nemotron-Labs-Diffusion: A Tri-Mode Language Model Unifying Autoregressive, Diffusion, and Self-Speculation Decoding https://megadose.ai/i/3ef202bc-76d8-4b8f-bfe4-c548f31b547a Megadose en 2026-07-07T00:00:00+00:00 RynnWorld-Teleop: An Action-Conditioned World Model for Digital Teleoperation https://megadose.ai/i/0b806aca-01b5-4ea5-b8be-ef498ae8fcb1 Megadose en 2026-07-07T00:00:00+00:00 RynnWorld-4D: 4D Embodied World Models for Robotic Manipulation https://megadose.ai/i/cfd09116-88e6-46dc-a271-ded534b07041 Megadose en 2026-07-07T00:00:00+00:00 From Foundation to Application: Improving VLA Models in Practice https://megadose.ai/i/692417ca-7518-4857-8190-314849227017 Megadose en 2026-07-07T00:00:00+00:00 Image2Sim: Scaling Embodied Navigation via Generative Neural Simulator https://megadose.ai/i/e8f813c7-960d-42ba-a90f-dd842e6c3150 Megadose en 2026-07-07T00:00:00+00:00 PluraMath: Extending Mathematical Reasoning Evaluation Beyond High-Resource Languages https://megadose.ai/i/9c386fbd-372b-431f-9dba-d7e6b68fb265 Megadose en 2026-07-07T00:00:00+00:00 TurnOPD: Making On-Policy Distillation Turn-Aware for Efficient Long-Horizon Agent Training https://megadose.ai/i/2cb5b18f-4f71-4602-a312-ad710b9a1060 Megadose en 2026-07-07T00:00:00+00:00 Vision as Unified Multimodal Generation https://megadose.ai/i/dd11e569-929e-40c2-840f-1936130f6229 Megadose en 2026-07-07T00:00:00+00:00 Quantifying and Expanding the Theoretical Capacity of Late-Interaction Retrieval Models https://megadose.ai/i/ada2aa0f-76cd-4912-89bb-615419f4048f Megadose en 2026-07-06T23:57:35+00:00 tencent/Hy3 https://megadose.ai/i/44c4d492-f757-4be3-a2c1-8c26a7c1bc53 Megadose en 2026-07-06T23:35:00+00:00 Syntiant, which develops low-power AI processors, files for a US IPO, reporting a $20.9M net loss on $64.5M in revenue for the three months ended March 31 (Pragyan Kalita/Reuters) https://megadose.ai/i/430e4a9b-4e45-4693-af6a-9fc983bab325 Megadose en 2026-07-06T23:21:03+00:00 US investors will soon get access to SK Hynix, another memory maker riding the AI boom https://megadose.ai/i/f87119db-9ef7-416c-8a11-5699dceb1fee Megadose en 2026-07-06T22:15:02+00:00 Anthropic researchers detail J-space, a small set of neural patterns in Claude that reveals internal thoughts that don't appear in the model's output (Anthropic) https://megadose.ai/i/3c6d42b7-c9a0-45bb-8ff8-5e97fdbf40f6 Megadose en 2026-07-06T17:59:59+00:00 From Fixed to Free Cameras: Calibration-Free View-Robust Vision-Language-Action Model https://megadose.ai/i/1c4a3478-0517-4875-8d05-e6c087868934 Megadose en 2026-07-06T17:59:58+00:00 Weak-to-Strong Generalization via Direct On-Policy Distillation https://megadose.ai/i/9b7dbbd0-65a1-4a7a-b947-3cbd2d8dd9b9 Megadose en 2026-07-06T17:59:35+00:00 LLM-as-a-Verifier: A General-Purpose Verification Framework https://megadose.ai/i/f8a72a16-5c6e-45ca-9b66-397137e0254c Megadose en 2026-07-06T17:56:12+00:00 Search Beyond What Can Be Taught: Evolving the Knowledge Boundary in Agentic Visual Generation https://megadose.ai/i/e4e6127d-8067-4edb-930f-9632dadd0635 Megadose en 2026-07-06T17:56:11+00:00 What Does a Discrete Diffusion Model Learn? https://megadose.ai/i/da5e3fae-3ce3-49f5-8cd1-62b50602dd42 Megadose en 2026-07-06T17:55:55+00:00 TabPack: Efficient Hyperparameter Ensembles for Tabular Deep Learning https://megadose.ai/i/922fb1e0-8731-4b41-80d1-2d4b1bd383c0 Megadose en 2026-07-06T17:55:12+00:00 CompactionRL: Reinforcement Learning with Context Compaction for Long-Horizon Agents https://megadose.ai/i/77c64b58-a70d-458f-a2d2-afe04d7a3c58 Megadose en 2026-07-06T17:55:05+00:00 Cortex: A Bidirectionally Aligned Embodied Agent Framework for Long-horizon Manipulation https://megadose.ai/i/0da1c217-572e-45fc-af49-6d887c9df2ba Megadose en 2026-07-06T17:53:32+00:00 Fitted Occupancy-Ratio Evaluation without Bellman Completeness https://megadose.ai/i/f94c7e9c-fbba-49b3-b5bd-ebc6be3962ee Megadose en 2026-07-06T17:47:31+00:00 GaP: A Graph-as-Policy Multi-Agent Self-Learning Harness For Variational Automation Tasks https://megadose.ai/i/818b9d02-aea6-4a64-8be8-367b53a8fe44 Megadose en 2026-07-06T17:42:59+00:00 SPEARBench: A Benchmark for Naturalness Evaluation in Streaming Speech-to-Speech Language Models https://megadose.ai/i/226585d4-d07f-411f-85b4-ac68e21eebaa Megadose en 2026-07-06T17:40:54+00:00 REDDIT: Correcting Model-Generated Timestamp Drift in ASR without Forgetting via Replay-Based Distribution Editing https://megadose.ai/i/ece33a64-5463-4b33-8677-259afc3ed728 Megadose en 2026-07-06T17:39:05+00:00 SovereignPA-Bench: Evaluating User-Owned Personal Agents under Evolving Intent, Platform Mediation, and Consent Constraints https://megadose.ai/i/dddce385-2bff-465f-8465-7be89254431f Megadose en 2026-07-06T17:36:28+00:00 Graph Sparse Sampling: Breaking the Curse of the Horizon in Continuous MDP Planning https://megadose.ai/i/1517abcf-90d9-4e6a-be23-85b5246081db Megadose en 2026-07-06T17:33:36+00:00 Faithfulness to Refusal: A Causal Audit of Neuron Selectors https://megadose.ai/i/197a6526-88c6-4cf1-8e21-48398dba7d74 Megadose en 2026-07-06T17:32:01+00:00 Selective Disclosure Watermarking for Large Language Models https://megadose.ai/i/ec066b12-9cc0-4f82-b18c-fa3cd412b4dc Megadose en 2026-07-06T17:31:52+00:00 Multiplayer Interactive World Models with Representation Autoencoders https://megadose.ai/i/cbf6cf7c-6f4c-4dd9-a908-a8539c386959 Megadose en 2026-07-06T17:27:16+00:00 OptiAgent: End-to-End Optimization Modeling via Multi-Agent Iterative Refinement https://megadose.ai/i/f3bb4788-231f-4e1d-a520-5cee0f3573e5 Megadose en 2026-07-06T17:21:16+00:00 TREK: Distill to Explore, Reinforce to Refine https://megadose.ai/i/1eaba0be-246e-4412-9b7e-7f0ea3802aae Megadose en 2026-07-06T17:01:51+00:00 Tencent releases Hy3, a 295B-parameter model that it says is competitive with GLM-5.1 and 5.2, under the Apache 2.0 license, following a preview launch in April (Sam Witteveen/VentureBeat) https://megadose.ai/i/0c94bef5-deca-47aa-9936-c83ec4bfb107 Megadose en 2026-07-06T16:58:05+00:00 Steering Optimisation Trajectories in Diffusion Representation Learning https://megadose.ai/i/3a355b77-68a1-4e34-a040-8dfe7cf94682 Megadose en 2026-07-06T16:56:04+00:00 How Much is Left? LLMs Linearly Encode Their Remaining Output Length https://megadose.ai/i/f418c51d-6087-460f-912a-40725037deab Megadose en 2026-07-06T16:49:10+00:00 Evaluating and Understanding Model Editing for Medical Vision Language Models https://megadose.ai/i/6eb241b2-5996-400c-b730-5ad2d65748af Megadose en 2026-07-06T16:47:44+00:00 OfficeCLI: Office suite for AI agents to read and edit Microsoft Office files https://megadose.ai/i/e9032936-d825-411b-abf6-e065b75ad1ea Megadose en 2026-07-06T16:43:04+00:00 Learning Only What Valid Adapters Can Express: Subspace-Constrained Adaptation Against Fine-Tuning Poisoning https://megadose.ai/i/ac3634c8-d4b3-4844-8ba3-92721a507db7 Megadose en 2026-07-06T16:40:23+00:00 MetaSkill-Evolve: Recursive Self-Improvement of LLM Agents via Two-Timescale Meta-Skill Evolution https://megadose.ai/i/33a38b1a-8ebb-459a-a188-e6bf492d7f45 Megadose en 2026-07-06T16:25:01+00:00 Source: Mercor hit over $2B in gross revenue run rate in June, double its pace in April, and is profitable on a free cash flow basis (Julia Hornstein/The Information) https://megadose.ai/i/d5cd17f4-429a-442b-880c-f6ff8f545808 Megadose en 2026-07-06T16:22:37+00:00 Untrusted Content Masking for Web Agents with Security Guarantees https://megadose.ai/i/92473b46-427d-4dec-8a71-4fb4d8a51536 Megadose en 2026-07-06T16:21:34+00:00 ProPS: Prompted Profile Synthesis for Natural Language-Conditioned Speaker Embedding Distributions https://megadose.ai/i/d9aaaa73-b59d-4b69-baca-db9118145b9f Megadose en 2026-07-06T16:20:19+00:00 Adaptive Inference Batching using Policy Gradients https://megadose.ai/i/4a953f60-7b25-4084-9876-5438c25821a9 Megadose en 2026-07-06T16:20:19+00:00 Target-Guided Selective Reweighting for Physics-Informed Neural Network Inverse Problems: A Transfer Learning Approach https://megadose.ai/i/fbf902ac-9650-429f-9bb8-0a82a1c92078 Megadose en 2026-07-06T16:18:05+00:00 Is the Geometry Doing the Work? An Operating-Point Audit of Hierarchy in Hyperbolic Vision-Language Models https://megadose.ai/i/c891c24a-0179-4722-898d-9ea092b2c9d8 Megadose en 2026-07-06T16:04:23+00:00 GeoFlow: Geo-Aware Modeling of Inter-Area Relationships in Origin-Destination Flow Prediction and Generation https://megadose.ai/i/067177fe-2f08-4260-9715-d1650a646b9c Megadose en 2026-07-06T15:59:18+00:00 FUSE: FK-Steered Multi-Modal Flow Matching for Efficient Simulation-Based Posterior Estimation https://megadose.ai/i/84682c49-f170-400c-adeb-7ef170ac2710 Megadose en 2026-07-06T15:59:17+00:00 Privacy-Preserving Robustness Verification for Neural Networks https://megadose.ai/i/8778c037-1a71-4b38-a5d0-f37bbc08a457 Megadose en 2026-07-06T15:57:49+00:00 Streaming Neural Speech Codecs through Time-Invariant Representations https://megadose.ai/i/090aa39a-7750-4f62-8516-bed07cad6b30 Megadose en 2026-07-06T15:50:44+00:00 Optimizing ML Workload Partitioning between CPUs and CIM Accelerators for Heterogeneous Computing https://megadose.ai/i/5f1ae33a-1ebd-4e47-92ef-2a98c5773d73 Megadose en 2026-07-06T15:47:37+00:00 MoP-JEPA: Hard-Assigned Predictor Mixtures for Stochastic JEPA World Models https://megadose.ai/i/791bf0aa-0b07-4fc5-af52-aa949fdd2999 Megadose en 2026-07-06T15:39:30+00:00 Progressive Refinement: An Iterative Pseudo-Labeling Approach for Mandarin-English Code-Switching ASR https://megadose.ai/i/ef219958-950b-4c22-acac-04ffe8d78cc3 Megadose en 2026-07-06T15:01:18+00:00 AMD Ryzen AI Halo – $4k AI Dev Kit https://megadose.ai/i/f075e770-d136-4476-b4eb-5ecdb5132986 Megadose en 2026-07-06T12:25:00+00:00 Anthropic signs a 20-year, $19B lease to use a TeraWulf data center in Kentucky, set to have a ~400MW capacity, with the first power delivery in H2 2027 (CNBC) https://megadose.ai/i/2d41e06b-ae9e-4b1a-b611-4e9fc8243c4a Megadose en 2026-07-06T12:15:22+00:00 AI agent training platform Bespoke Labs raised $40M across a seed led by 8VC and a Series A led by Wing, saying the funding gives it almost two years of runway (Natalie Breymeyer/Axios) https://megadose.ai/i/9dc183bc-a604-4df4-998e-b6acf891c773 Megadose en 2026-07-06T11:50:01+00:00 Filing: Brookfield-backed data center company Csquare seeks to raise up to $1.35B in an IPO, selling 50M shares at $23 to $27 each for an up to $4.18B valuation (Ryan Vlastelica/Bloomberg)