Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
A new arXiv paper (arXiv:2609.05459v1) proposes a deterministic framework that compiles games specified in the Video Game Description Language (VGDL) into Dynamic Structural Causal Models, directly translating sprite dyn…
Researchers introduced EnvCraft, an automated framework that synthesizes executable environments and training data for Agentic Reinforcement Learning, detailed in arXiv paper 2609.05576v1. EnvCraft's environment synthesi…
Researchers studying Looped GPT-BERT in the BabyLM 2026 Strict-small setting trained a 12.18M-parameter model on a preprocessed 7.48M-word English corpus, using four physical layers for twelve recurrent traversals, and t…
A new arXiv paper (2609.05448v1) formulates structured post-training pruning of vision and language transformers as a damage-aware multi-armed bandit problem under a fixed candidate-evaluation budget, masking attention h…
A reasoning-aware compression framework that selectively restores the most quantization-sensitive circuits to FP16 achieves Pareto-optimal energy-accuracy points unreachable by uniform quantization, according to an arXiv…
Researchers proposed TEAM-Design, a rule that assigns every task two replay probabilities — one per baseline — to decide under a fixed replay budget whether a human-AI workflow should be kept over the human alone or the …
Researchers introduced Capsule Lens, a framework that fits a simple geometric "capsule" defined by interpretable parameters in closed form to each concept's region in a machine learning model's internal representations, …
Researchers proposed NOMAD-RL (Neural Online Meta-Adaptation for Dynamics), a general-purpose reinforcement learning controller that transfers across heterogeneous thermal zones through a universal, non-invasive thermost…
A reinforcement learning framework that drives exploration through intrinsic curiosity achieved competitive performance against Proximal Policy Optimization (PPO) and the Intrinsic Curiosity Module (ICM) on the LunarLand…
A new arXiv paper (2609.05688v1) connects denoising score matching to maximum likelihood and Expectation-Maximization in mixed linear regression with unknown mixing weights. The analysis shows that, under mild regularity…
Researchers introduced GraphNOSE, an open-source graph transformer framework that predicts multi-label odor descriptors from SMILES strings for single molecules and binary mixtures, according to an arXiv paper (2609.0569…
A new arXiv paper (2609.05727v1) introduces Newton Matching, a unified framework that treats fine-tuning and sampling in generative modeling as iterative optimization over canonical models rather than isolated losses. Th…
A hybrid candidate-generation architecture combining item-based collaborative filtering with graph neural network retrieval improved Recall@300 by 14.8% over the strongest baseline on a vacation rental platform with over…
Researchers released Video-MOPD-8B, an open-weight video understanding model trained via Multi-Teacher On-Policy Distillation (MOPD), which the paper reports achieves state-of-the-art performance among models of comparab…
Researchers introduced VANTAGE-Bench, a benchmark measuring an "Infrastructure AI Gap" in Vision-Language Models (VLMs) that rely on fixed cameras for open-loop tasks such as safety monitoring and operational logging. Ev…
A new arXiv paper (2609.09462v1) reports that factorizing the dense CoOp prompt matrix P ∈ R^(m×d) as P = BA cuts trainable prompt parameters from md to r(m+d), and to rd once the token-side factor B is fixed, matching o…
A new deep, global Structure-from-Motion framework learns camera poses from noisy view-graphs using a permutation-equivariant, edge-conditioned graph neural network that outputs globally consistent camera extrinsics with…
Researchers proposed DrugReason, a multi-view reasoning framework that integrates knowledge graph-based reasoning with LLM-generated mechanistic inference for drug repurposing, according to an arXiv paper (arXiv:2609.067…
Researchers introduced PCSDiff, a cascaded task-decoupled diffusion framework for 10-day precipitation bias correction and downscaling, according to an arXiv paper (arXiv:2609.06942v1). Evaluated against CMA-CRA observat…
A projection-based, black-box method operating directly on output embeddings recovers meaningful chronological and spatial structure from language-model representations, according to arXiv paper 2609.05721v1. The method …