TechniqueRLHF / Alignment8 recent entries8 Oct 2026GRACE: Generation-aware latent compression for efficient video generationarXiv:2610.10524v1 Announce Type: new Abstract: Highly compressed video autoencoders offer an effective way to accelerate video diffusion models, as the Diffusion Transformer (DiT) operates on far few→8 Oct 2026Event-Aligned Visual Action Reasoning for World Action ModelsarXiv:2610.09427v1 Announce Type: new Abstract: World-Action Models (WAMs) utilize future visual prediction as an intermediate reasoning process to guide action generation. However, existing WAMs typi
TechniqueRAG8 recent entries2 Oct 2026Exploring More, Reasoning Better: Stepwise Risk-Sensitive GRPO for Diffusion Language ModelsarXiv:2610.00661v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) generate text by denoising a sequence or successive blocks, allowing several tokens to be revealed in parallel→5 Oct 2026JOVE: Joint Execution and Verification for Resource-Aware LLM Task GraphsarXiv:2610.03296v1 Announce Type: new Abstract: Complex reasoning queries can be decomposed into directed acyclic task graphs and distributed across heterogeneous LLMs, reducing latency through parall→5 Oct 2026How To Train Your World Model: Fine-tuning vs RAG for LM-based World ModelingarXiv:2610.02542v1 Announce Type: new Abstract: World models (WMs) simulate the transition dynamics of environments, enabling agents to plan over the consequences of their actions. In text-based envir→7 Oct 2026Rethinking access control for RAG with Amazon Quick and Amazon BedrockEnterprise RAG unlocks insights from knowledge sources like SharePoint, Google Drive, and Confluence, but those sources carry complex permissions. Learn how Amazon Quick and Amazon Bedrock Knowledge B→8 Oct 2026Robotic Ultra-Long-Horizon Manipulation Skills via Human-guided Lifelong Code GenerationarXiv:2509.18597v3 Announce Type: replace Abstract: Large language models (LLMs) can translate natural-language instructions for robotic manipulation into executable code, but ambiguity, noisy generat→8 Oct 2026Mechanics of Long-Context Hybrid Models Part 1.1: From Hybrid Attention to Hybrid PositionarXiv:2610.10114v1 Announce Type: new Abstract: The architectural design of Large Language Models (LLMs) is shifting from traditional full-attention-only models to hybrid models, which combine differe→8 Oct 2026HealthcareNLP: where are we and what is next?arXiv:2512.08617v2 Announce Type: replace Abstract: This tutorial focused on Healthcare Domain Applications of NLP, what we have achieved around HealthcareNLP, and the challenges that lie ahead for th→12 Oct 2026BoundaryEvo: Data-Model Co-Evolution at the Capability Boundary for Long-Chain Reasoning DistillationarXiv:2610.13285v1 Announce Type: new Abstract: Distilling long-chain reasoning trajectories from large teacher models provides an effective way to improve the reasoning ability of small language mode
TechniqueAgents8 recent entries6 Oct 2026Researching from scratch! [R]Hey everyone! I am trying to go really deep into Machine Learning and Artificial Intelligence. Not just learning how to use the latest models and frameworks, but understanding how we reached here from→7 Oct 2026ROMA: LLM System for Real-World Object-Centric Multi-Sensory Active PerceptionarXiv:2610.06955v1 Announce Type: cross Abstract: Humans inherently understand the physical world through an active process. When sensory evidence is insufficient to infer physical properties, we natu→7 Oct 2026Learning Scientific Exploration from Human Research Decision TrajectoriesarXiv:2610.07184v1 Announce Type: cross Abstract: A key challenge in building AI systems for scientific research is enabling extit{scientific exploration}: the systematic process of investigating unkn→7 Oct 2026Learning Explainable Representations of Complex Game-playing StrategiesarXiv:2610.07638v1 Announce Type: new Abstract: As part of learning to play complex games, human players develop develop abstractions for concepts and strategies of gameplay consistent with game rules→8 Oct 2026Evaluating the Transfer of Co-Evolved Communication from 2D to 3D SimulationarXiv:2610.09280v1 Announce Type: new Abstract: This work examines the transfer of a co-evolved communication mechanism between two robotic agents from a discrete two-dimensional (2D) simulator to a t→9 Oct 2026Skill-V: Verifiable Self-Evolving Skill Library for Interactive AgentsarXiv:2610.11781v1 Announce Type: new Abstract: Interactive agents can turn experience into reusable skills, yet existing self-evolving skill libraries primarily improve by accumulating new knowledge.→9 Oct 2026Curating Always-Loaded Context for LLM Agents: A Capacitated Assortment Model with Censored FeedbackarXiv:2610.11007v1 Announce Type: new Abstract: At the start of every session, LLM agents load a fixed context file, such as exttt{AGENTS.md}. Each loaded token in the file is charged again in every l→9 Oct 2026AgentGarten: Code Worlds for Evolving AgentsarXiv:2610.12374v1 Announce Type: new Abstract: Interactive virtual worlds allow agents to learn through exploration and interaction. What agents can learn is bounded by the environments they practice
TechniqueFine-tuning8 recent entries8 Oct 2026GRACE: Generation-aware latent compression for efficient video generationarXiv:2610.10524v1 Announce Type: new Abstract: Highly compressed video autoencoders offer an effective way to accelerate video diffusion models, as the Diffusion Transformer (DiT) operates on far few→9 Oct 2026Prosody-to-Text: Predicting text from low-pass filtered speecharXiv:2610.11544v1 Announce Type: new Abstract: While predicting prosody from text is an established task in the field, the opposite direction, predicting text that fits a given prosodic pattern, rema→9 Oct 2026Parametric Trajectory Distillation for Few-Step Video GenerationarXiv:2610.11498v1 Announce Type: new Abstract: Video diffusion and flow models require many sequential evaluations, making generation computationally expensive. Few-step distillation reduces this cos→9 Oct 2026Controllable Exaggeration for Generative Motion Models via Training-Time Adaptation and Inference-Time GuidancearXiv:2610.12316v1 Announce Type: new Abstract: Recent motion generative models have demonstrated strong capabilities in synthesizing physically plausible character motion, but often overlook establis→9 Oct 2026Can a System-One LLM Perform Knowledge Tracing When Few or No Learners Are Logged?arXiv:2610.11135v1 Announce Type: new Abstract: Knowledge tracing (KT) models need many logged learners, so a new course or platform starts without a usable model. In LLM-based KT the LLM generates th→12 Oct 2026When Generative Models Act as SimulatorsarXiv:2610.12847v1 Announce Type: cross Abstract: A simulator summarizes an environment in a state that users can read and change. Generative models learn representations of the world without explicit→12 Oct 2026One Thing Leads to Another: The Essentials of Time Series Causal DiscoveryarXiv:2610.13430v1 Announce Type: cross Abstract: Time series causal discovery --- the process of detecting causation between time series --- has its roots in signal processing research from over a ha→12 Oct 2026Low-Rank Kolmogorov-Arnold Generator Flows for Interpretable Operator LearningarXiv:2610.13028v1 Announce Type: new Abstract: Neural operators that learn physical laws from data should generalize beyond their training distribution and expose the learned law in closed form. The
TechniqueMultimodal8 recent entries9 Oct 2026Softmax Attention on Gaussian Mixtures: Linear When It Can, Selective When It MustarXiv:2610.11798v1 Announce Type: cross Abstract: Softmax attention, at the heart of Transformers, has demonstrated remarkable capabilities. Yet its underlying mechanisms remain only partially underst→9 Oct 2026OuroWorld: Bringing Any 3D World Alive as Diverse, Endlessly Looping 3D CinemagraphsarXiv:2610.12461v1 Announce Type: new Abstract: Recent 3D world models generate photorealistic, explorable scenes that remain frozen in time. OuroWorld is a mask-free framework that turns any static 3→9 Oct 2026Omni-Diffusion-Distill: Few-Step Distillation of Unified Multimodal Diffusion Large Language ModelsarXiv:2610.10990v1 Announce Type: new Abstract: Unified multimodal diffusion large language models (dLLMs) offer a single architecture for both image generation and multimodal understanding, but their→9 Oct 2026Mid-Training Language Models on Raw VideoarXiv:2610.11019v1 Announce Type: cross Abstract: Multimodal large language models learn mostly from paired image-text data or annotated video, and raw web video is rarely used to further train an exi→9 Oct 2026ImagiNav: Scalable Embodied Navigation via Generative Visual Prediction and Inverse DynamicsarXiv:2603.13833v2 Announce Type: replace Abstract: Enabling robots to navigate open-world environments via natural language is critical for general-purpose autonomy. Yet, Vision-Language Navigation h→9 Oct 2026Evi-VN: Hard Region Guided Virtual Node Evidence Injection for GNN-Based Fraud DetectionarXiv:2610.11665v1 Announce Type: new Abstract: Online platforms contain growing numbers of bots, deceptive reviewers, and scam accounts that imitate legitimate users. Such camouflage blurs graph neig→9 Oct 2026Beyond Visual Enhancement: Adaptive Multi-Context Steering to Mitigate LVLM HallucinationsarXiv:2610.11907v1 Announce Type: cross Abstract: Hallucination remains a significant challenge in Large Vision-Language Models (LVLMs). Existing training-free methods generally mitigate hallucination→12 Oct 2026SpatialASM: Internalizing Active Mentalization for Spatial ReasoningarXiv:2610.13832v1 Announce Type: cross Abstract: When solving complex spatial reasoning tasks, ordinary humans naturally revisit the scenes from task-relevant viewpoints and mentally reconstruct the
TechniqueSafety8 recent entries10 Jun 2026There is a lot of justified anger at Anthropic for sandbagging Fable 5 for AI development tasks. But an unanticipated side effect is that th…There is a lot of justified anger at Anthropic for sandbagging Fable 5 for AI development tasks. But an unanticipated side effect is that third-party evaluators can no longer credibly use the model fo→29 Jul 2026Self-host a validated coding assistant with NeMo GuardrailsNVIDIA tutorial: StarCoder2-7B NIM behind NeMo Guardrails, plus CI checks and commit trailers so generated patches stay auditable.→17 Aug 2026We’ve got a workshop on production retrieval-augmented generation with open models, benchmarked end to end, thought it’d be relevant here [D]There’s a hands-on workshop on August 29 that builds and benchmarks this properly, end to end, using entirely open models, no API calls involved. Led by Ben Auffarth, AI Consultant and Founder of Chel→8 Sept 2026Govern models with MLflow and Amazon SageMaker AI Model Registry sync: Part 1Managed MLflow on Amazon SageMaker AI now syncs richer model metadata (training metrics, evaluation results, inference specs, and lineage) into the SageMaker AI Model Registry, with lifecycle stage pr→9 Sept 2026Beyond embedding: How to secure AI/BI Dashboards for every viewerAn embedded AI/BI dashboard can be securely shared with multiple viewers while ensuring that each viewer can see only their authorized rows. This protection is achieved through group‑based row‑level s→25 Sept 2026Guardrails or Roadblocks? Effects of Pedagogical Style and Context Awareness in AI Teaching Assistants for ProgrammingarXiv:2609.29995v1 Announce Type: cross Abstract: AI teaching assistants (AI TAs) backed by large language models (LLMs) and pedagogical guardrails are increasingly being integrated into programming c→30 Sept 2026Query claims in natural language with Amazon Bedrock Knowledge BasesThis technical how-to builds a conversational claims assistant on Amazon Bedrock Knowledge Bases that answers natural-language questions with citations. It covers ingesting claim documents from Amazon→7 Oct 2026Learn more about identifying AI-generated content: https://blog.google/innovation-and-ai/models-and-research/google-deepmind/synth-id-ai-con…Google AI has published a new blog post titled “Learn more about identifying AI‑generated content,” announcing tools that make it easier to spot machine‑created text on a global scale. The article – r