Audio-Visual Continual Test-Time Adaptation without Forgetting
arXiv:2602.18528v1 Announce Type: new Abstract: Audio-visual continual test-time adaptation involves continually adapting a source audio-visual model at test-time, to unlabeled non-stationary domains, where either or both modalities can be distributionally shifted, which hampers online cross-modal learning and eventually leads to...
Sub-City Real Estate Price Index Forecasting at Weekly Horizons Using Satellite Radar and News Sentiment
arXiv:2602.18572v1 Announce Type: new Abstract: Reliable real estate price indicators are typically published at city level and low frequency, limiting their use for neighborhood-scale monitoring and long-horizon planning. We study whether sub-city price indices can be forecasted at weekly frequency...
GIST: Targeted Data Selection for Instruction Tuning via Coupled Optimization Geometry
arXiv:2602.18584v1 Announce Type: new Abstract: Targeted data selection has emerged as a crucial paradigm for efficient instruction tuning, aiming to identify a small yet influential subset of training examples for a specific target task. In practice, influence is often measured...
Ensemble Prediction of Task Affinity for Efficient Multi-Task Learning
arXiv:2602.18591v1 Announce Type: new Abstract: A fundamental problem in multi-task learning (MTL) is identifying groups of tasks that should be learned together. Since training MTL models for all possible combinations of tasks is prohibitively expensive for large task sets, a...
Diagnosing LLM Reranker Behavior Under Fixed Evidence Pools
arXiv:2602.18613v1 Announce Type: new Abstract: Standard reranking evaluations study how a reranker orders candidates returned by an upstream retriever. This setup couples ranking behavior with retrieval quality, so differences in output cannot be attributed to the ranking policy alone. We...
Learning Invariant Visual Representations for Planning with Joint-Embedding Predictive World Models
arXiv:2602.18639v1 Announce Type: new Abstract: World models learned from high-dimensional visual observations allow agents to make decisions and plan directly in latent space, avoiding pixel-level reconstruction. However, recent latent predictive architectures (JEPAs), including the DINO world model (DINO-WM), display a...
Adaptive Time Series Reasoning via Segment Selection
arXiv:2602.18645v1 Announce Type: new Abstract: Time series reasoning tasks often start with a natural language question and require targeted analysis of a time series. Evidence may span the full series or appear in a few short intervals, so the model...
Information-Guided Noise Allocation for Efficient Diffusion Training
arXiv:2602.18647v1 Announce Type: new Abstract: Training diffusion models typically relies on manually tuned noise schedules, which can waste computation on weakly informative noise regions and limit transfer across datasets, resolutions, and representations. We revisit noise schedule allocation through an information-theoretic...
Large Causal Models for Temporal Causal Discovery
arXiv:2602.18662v1 Announce Type: new Abstract: Causal discovery for both cross-sectional and temporal data has traditionally followed a dataset-specific paradigm, where a new model is fitted for each individual dataset. Such an approach limits the potential of multi-dataset pretraining. The concept...
In-Context Planning with Latent Temporal Abstractions
arXiv:2602.18694v1 Announce Type: new Abstract: Planning-based reinforcement learning for continuous control is bottlenecked by two practical issues: planning at primitive time scales leads to prohibitive branching and long horizons, while real environments are frequently partially observable and exhibit regime shifts...
Insertion Based Sequence Generation with Learnable Order Dynamics
arXiv:2602.18695v1 Announce Type: new Abstract: In many domains generating variable length sequences through insertions provides greater flexibility over autoregressive models. However, the action space of insertion models is much larger than that of autoregressive models (ARMs) making the learning challenging....
Phase-Consistent Magnetic Spectral Learning for Multi-View Clustering
arXiv:2602.18728v1 Announce Type: new Abstract: Unsupervised multi-view clustering (MVC) aims to partition data into meaningful groups by leveraging complementary information from multiple views without labels, yet a central challenge is to obtain a reliable shared structural signal to guide representation...
When World Models Dream Wrong: Physical-Conditioned Adversarial Attacks against World Models
arXiv:2602.18739v1 Announce Type: new Abstract: Generative world models (WMs) are increasingly used to synthesize controllable, sensor-conditioned driving videos, yet their reliance on physical priors exposes novel attack surfaces. In this paper, we present Physical-Conditioned World Model Attack (PhysCond-WMA), the first...
From Few-Shot to Zero-Shot: Towards Generalist Graph Anomaly Detection
arXiv:2602.18793v1 Announce Type: new Abstract: Graph anomaly detection (GAD) is critical for identifying abnormal nodes in graph-structured data from diverse domains, including cybersecurity and social networks. The existing GAD methods often focus on the learning paradigms of "one-model-for-one-dataset", requiring dataset-specific...
Vectorized Bayesian Inference for Latent Dirichlet-Tree Allocation
arXiv:2602.18795v1 Announce Type: new Abstract: Latent Dirichlet Allocation (LDA) is a foundational model for discovering latent thematic structure in discrete data, but its Dirichlet prior cannot represent the rich correlations and hierarchical relationships often present among topics. We introduce the...
SGNO: Spectral Generator Neural Operators for Stable Long Horizon PDE Rollouts
arXiv:2602.18801v1 Announce Type: new Abstract: Neural operators provide fast PDE surrogates and often generalize across parameters and resolutions. However, in the short train long test setting, autoregressive rollouts can become unstable. This typically happens for two reasons: one step errors...
Exact Attention Sensitivity and the Geometry of Transformer Stability
arXiv:2602.18849v1 Announce Type: new Abstract: Despite powering modern AI, transformers remain mysteriously brittle to train. We develop a stability theory that explains why pre-LayerNorm works, why DeepNorm uses $N^{-1/4}$ scaling, and why warmup is necessary, all from first principles. Our...
Rank-Aware Spectral Bounds on Attention Logits for Stable Low-Precision Training
arXiv:2602.18851v1 Announce Type: new Abstract: Attention scores in transformers are bilinear forms $S_{ij} = x_i^\top M x_j / \sqrt{d_h}$ whose maximum magnitude governs overflow risk in low-precision training. We derive a \emph{rank-aware concentration inequality}: when the interaction matrix $M =...
PCA-VAE: Differentiable Subspace Quantization without Codebook Collapse
arXiv:2602.18904v1 Announce Type: new Abstract: Vector-quantized autoencoders deliver high-fidelity latents but suffer inherent flaws: the quantizer is non-differentiable, requires straight-through hacks, and is prone to collapse. We address these issues at the root by replacing VQ with a simple, principled,...
Court holds that U.S. Postal Service can’t be sued over intentionally misdelivered mail
A divided Supreme Court sided with the federal government on Tuesday in U.S. Postal Service v. Konan, a dispute over mishandled mail. Writing for a 5-4 majority, Justice Clarence Thomas […]The postCourt holds that U.S. Postal Service can’t be sued...
In Defense of Substantive Due Process
Introduction Originalism has a branding and substance problem.[1] If originalism is what it purports to be—impartial and value-free enforcement of the Founders’ intention and “the only approach to text that is compatible with democracy”[2]—more Americans would have faith in the...
Chill
Introduction No concept is more pervasive in the law of freedom of speech than chill.[1] The chilled speech doctrine guards against self-censorship: it permits First Amendment challenges based on the allegation that a law deters the plaintiff or others from...
DJI sues the FCC for “carelessly” restricting its drones
DJI lawsuit says company has been "severely harmed by the FCC’s ruling."
UK fines Reddit for not checking user ages aggressively enough
UK agency alleges "Reddit failed to apply any robust age assurance mechanism."
Nvidia challenger AI chip startup MatX raised $500M
The startup was founded by former Google TPU engineers in 2023.
Uber engineers built an AI version of their boss
Uber CEO Dara Khosrowshahi said the company’s employees have gone all in on AI, going so far as to build a chatbot of him that they use to practice their pitches.
Anthropic won’t budge as Pentagon escalates AI dispute
The Pentagon has given Anthropic until Friday to loosen AI guardrails or face potential penalties, escalating a high-stakes dispute that raises questions about government leverage, vendor dependence, and investor confidence in defense tech.
Google adds a way to create automated workflows to Opal
The company said that a new agent being introduced in Opal will allow users to create mini-apps that can let them plan and execute tasks using text prompts.
OpenAI COO says ‘we have not yet really seen AI penetrate enterprise business processes’
There is a lot of talk around AI agents taking over business processes and claiming that "SaaS is dead." While these predictions have moved SaaS stocks at times, they haven't really come true.
Music generator ProducerAI joins Google Labs
Wyclef Jean used Google's AI music tools on his new song "Back in Abu Dhabi."