NVIDIA Releases Nemotron-Labs-3-Puzzle-75B-A9B: A Compressed Hybrid MoE LLM Delivering 2.03x Server Throughput at Matched User Throughput
NVIDIA has released Nemotron-Labs-3-Puzzle-75B-A9B, a compressed variant of Nemotron-3-Super. Iterative Puzzle alternates hardware-aware structural compression with short knowledge distillation recovery phases. The model drops from 120.7B total / 12.8B active parameters to 75.3B / 9.3B. On a single ...
Google AI Studio Adds Import from GitHub to Build a Deployable App
Google AI Studio is rolling out Import from GitHub in Build mode. It transforms an existing repo into a runtime-compatible format. You can then iterate on it, deploy it, and more.
The post Google AI Studio Adds Import from GitHub to Build a Deployable App appeared first on MarkTechPost.
TriRoute: Unified Learned Routing for Joint Adaptive Attention, Experts, and KV-Cache Allocation
arXiv:2607.06601v1 Announce Type: new
Abstract: Conditional computation can decouple language model quality from per-token inference cost, yet leading techniques act on a single axis in isolation: Mixture-of-Experts (MoE) sparsifies the FFN, Mixture-of-Depths (MoD) skips whole transformer blocks, a...
A Quiet Failure in Calibrated Virtual Screening: Marginal Conformal Prediction Under-Covers the Minority Class, and a Class-Conditional Fix Recovers It
arXiv:2607.06605v1 Announce Type: new
Abstract: Conformal prediction is being adopted in drug discovery to put an honest number on model reliability: pick an error rate alpha, and the method returns prediction sets containing the true label with probability at least 1 - alpha. We show this guarante...
D2PO: Optimizing Diffusion Samplers via Dynamic Preference
arXiv:2607.06609v1 Announce Type: new
Abstract: We propose D2PO (Dynamic Direct Preference Optimization), a principled framework for optimizing diffusion sampling policies with respect to timestep schedules and classifier-free guidance (CFG) weights. Our work is motivated by a fundamental limitatio...
When Does In-Context Search Help? A Sampling-Complexity Theory of Reflection-Driven Reasoning
arXiv:2607.06720v1 Announce Type: new
Abstract: Training large language models (LLMs) with extended reasoning has enabled in-context search, in which models iteratively generate, critique, and revise solution attempts. We provide a theoretical analysis of in-context search by modeling it as approxi...
arXiv:2607.06757v1 Announce Type: new
Abstract: Agent-based modeling (ABM) has the capability to model millions of individuals and their interactions, which is useful for policy making. However, ABMs have traditionally relied on static prior, which prevents the models from adapting to real-time cha...
QANTIS: Hardware-Calibrated Sequential POMDP Belief Updates on IBM Heron
arXiv:2607.06760v1 Announce Type: new
Abstract: Autonomous systems under partial observability act on beliefs, not raw sensor events. QANTIS treats the quantum processor as a calibrated belief-update service in that loop: it receives a prior and an observation model, estimates the rare-event eviden...
Cost-Effective Agent Harnesses for Abstract Reasoning and Generalization on ARC-AGI-1
arXiv:2607.06764v1 Announce Type: new
Abstract: Recent progress on ARC-AGI-1 from disclosed architectures has come broadly from two regimes: heavy test-time compute over frontier models (evolutionary search, exhaustive sampling, extended chain-of-thought), or benchmark-specific training in which sm...
Robbyant Releases LingBot-VLA 2.0: An Open-Source 6B Vision-Language-Action (VLA) Model for Cross-Embodiment Robot Manipulation
Ant Group's Robbyant has released LingBot-VLA 2.0, an Apache-2.0 vision-language-action model for cross-embodiment robot manipulation. The 6B checkpoint is pretrained on roughly 60,000 hours of data, spanning 50,000 hours of robot trajectories across 20 robot configurations and 10,000 hours of egoce...
AI Weekly Issue #512: Robotics Is Moving Fast: IPOs, New Models, and Smarter Robots
Three humanoid companies moved toward the public markets in a single week. Agility filed to go public via SPAC at $2.5 billion, Unitree cleared its Shanghai IPO, and Tesla started turning the line that built its last Model S into an Optimus factory. Mistral shipped a robot brain that finds its way w...
Incentivizing Temporal-Awareness in Egocentric Video Understanding Models
Multimodal large language models (MLLMs) have recently shown strong performance in visual understanding, yet they often lack temporal awareness, particularly in egocentric settings where reasoning depends on the correct ordering and evolution of events. This deficiency stems in part from training ob...
Biomni can analyze mountains of medical data, spot patterns humans might miss, and even design experiments—helping researchers make discoveries faster in the race to cure disease.
Recursive Language Models Meet Uncertainty: The Surprising Effectiveness of Self-Reflective Program Search for Long Context
Long-context handling remains a core challenge for language models: even with extended context windows, models often fail to reliably extract, reason over, and use the information across long contexts. Recent works like Recursive Language Models (RLMs) have approached this challenge by agentic way o...
SpaceXAI Releases Grok 4.5, a Cursor-Trained Model for Coding, Agentic Tasks, and Knowledge Work at $2/M Input
SpaceXAI released Grok 4.5, a Cursor-trained model for coding, agentic tasks, and knowledge work. It serves at 80 TPS, costs $2/$6 per million tokens, and ranks #1 on Harvey's Legal Agent Benchmark.
The post SpaceXAI Releases Grok 4.5, a Cursor-Trained Model for Coding, Agentic Tasks, and Knowledge ...
SpaceXAI releases Grok 4.5, which Elon describes as an ‘Opus-class model’
Elon Musk's tech company released the newest version of Grok on Wednesday, promising a cheaper, more efficient alternative to other powerful AI models.
This startup thinks robotics is about to have its ChatGPT moment
General Intuition is betting millions of hours of video game data can train the foundation models for physical AI, making it easier to build smarter robots with minimal real-world data.
AI Reality Check: The Rise of Algorithmic Middle Management
AI is quietly absorbing the traditional functions of middle management — monitoring, prioritization, coordination, and evaluation. Week 20 of AI Reality Check explores how algorithmic systems are reshaping organizational power, worker autonomy, and the future of corporate structure.
Google AI Studio Adds ‘Import from GitHub’ to Build Mode, Turning an Existing Repo Into an Editable, Deployable App
Google AI Studio is rolling out Import from GitHub in Build mode. It transforms an existing repo into a runtime-compatible format. You can then iterate on it, deploy it, and more. Here is what changes for developers.
The post Google AI Studio Adds ‘Import from GitHub’ to Build Mode, Turning an Exist...
The feature can do things like apply cinematic relighting to brighten up a dark clip, swap out a plain background for something fun, or add artistic styles to videos.
DeepSeek DSpark: The Speculative Decoding Trick Behind 400% Faster LLM
DeepSeek’s new DSpark module brings speculative decoding to DeepSeek-V4. It might look like a niche inference tweak, but in production it boosted per-user generation speed by 60 to 85 percent with no drop in model quality. What sets DSpark apart is that it tackles two longstanding problems at once, ...
OpenAI Releases GPT-Live and GPT-Live-1 mini: Full-Duplex Voice Models That Delegate Deeper Reasoning to GPT-5.5
OpenAI released GPT-Live, a new generation of voice models now powering ChatGPT Voice. It uses a full-duplex architecture to listen and speak at once, and delegates search and reasoning to GPT-5.5.
The post OpenAI Releases GPT-Live and GPT-Live-1 mini: Full-Duplex Voice Models That Delegate Deeper R...