The Archive

XGenAct: world action model for robot manipulation integrating RGB, depth, surface normals, and functional role prediction via cross-task generation.

Extended kernel SVD method for joint nonlinear feature learning across multiple asymmetric data sources.

ParaGeo decomposes paralinguistic variation in speech via frozen LM K/V projections; 80-control probe on GLM-4-Voice.

MIRTO standardizes evaluation choices for unsupervised brain MRI anomaly detection with registration-gated validation and explicit reporting of thresholds and metrics.

We explore catastrophic forgetting in the context of large pre-trained models. By considering forgetting as a geometric problem in the input space of each weight matrix, we uncover a natural retention objective under which updates produced by gradient-based optimizers are suboptimal. Following this observation, we propose Local Support Learning (LSL), a general-purpose framework that augments gradient-based training for retention of prior capabilities without access to prior data. During a new learning phase, LSL pairs two components with distinct roles: a standard weight adapter, trained as ...

GeoLatent decomposes spatial latents for 3D reasoning in vision-language models using geometry-guided optimization to stabilize directional representations.

Discrete Wasserstein geometry framework for one-step generative modeling on finite state spaces with amortized training.

Stylometric analysis of 137k French Compar:IA votes reveals formatting and lexical diversity influence LLM rankings beyond content quality.

CAS II develops algorithmic statistics over symmetric group-orbit partitions using Kolmogorov structure functions.

Instagram is the latest social media platform to add built-in AI-powered features that will give users feedback on their posts. The company announced Wednesday that its standalone Edits app will now include an AI "creative assistant" that pulls in data from a user's Instagram account and offers suggestions on what to change. Using the new assistant, creators can ask open-ended questions about what's trending on Instagram, why one video worked and another didn't, or what piece of content resonated most with their audience. The tool pulls in data and metrics such as likes, views, retention on v...

Token-level reward signals using visual dependency and entropy metrics improve multimodal reasoning in LLMs via reinforcement learning.

Parameter symmetries in overparameterized networks decouple function from representational geometry; implications for inference.

Vision-language models have made it possible to build visual AI agents that understand video at production scale. The harder problem is turning that capability... Vision-language models have made it possible to build visual AI agents that understand video at production scale. The harder problem is turning that capability into a maintainable system that combines ingestion, stream processing, event detection, retrieval, summarization, and reporting. The NVIDIA Metropolis Blueprint for Video Search and Summarization (VSS) and its agent skills help… Source

FurE reconstructs animal fur geometry from multi-view images using latent fields and PCA decoding, addressing fine-scale detail without fur datasets.

CoSE-E benchmark evaluates code-switched ASR in enterprise voice agents across 5 language pairs with task-failure propagation metrics.

GPUPhysBench: 50-task benchmark for coding agents on GPU physics simulation covering fluids, solids, and granular materials with efficiency metrics.

Mechanistic analysis reveals read-write asymmetry in Transformer attention/FFN layers: blocks ignore extreme-value features while reading but not writing.

Natural Boundary Learning applies Stein contrasts to contextual bandits with monotone link functions, avoiding reward function estimation.

State-space model for metric 3D point tracking from monocular video, enabling robot navigation and autonomous driving decisions in absolute meters rather than pixels.

Permutation-equivariant flow matching generates neural network weights across tasks while handling hidden unit permutation symmetry.

Clinical LLMs exhibit faithfulness gaps: high accuracy without correct reasoning. Introduces ESI, CFS, PSS metrics for explanation fidelity.

Re-derivability metric predicts cost of upstream faults in staged LLM agent pipelines; inspector agents grounded in original problem show +0.6 recovery gain.

Empirical study identifies design choices for viewpoint-generalizable visuomotor policies: dense visual tokens and geometric reasoning in action heads improve robustness.

OC-GS refines camera angles and geometry for 3D reconstruction from sparse, irregularly-spaced turntable views.

Flow matching framework unifies neural representational dissimilarity metrics for comparing neural codes across domains.

Theoretical analysis of transformer in-context learning under heterogeneous geometric data with minimax-optimal bounds and geometry adaptation.

Study reveals KV cache reuse evaluation metrics artificially inflate effectiveness; proposes faithful accuracy measurement methodology for retrieval-augmented generation.

Datasette 1.0a41 adds OpenTelemetry support and refactors modals to Web Component.

MISVO steers frozen LMs at test-time via constrained logit intervention, using KL geometry to minimize output distribution shift.

Study across 13 MLLMs shows standard similarity metrics misattribute visual-language alignment; noise-injection exposes shallow integration.