Search: “pretraining”

Steerspeech: Activation Steering For Emotion Control In Generated Speech

arXiv:2610.10415v1 Announce Type: cross Abstract: Pretrained text-to-speech (TTS) models can generate expressive speech, but reliable inference-time emotion control remains challenging: prompts and re

8 Oct 2026

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Thinking in Depth: Retrospective Inference for Tabular Foundation Models

DGX agent

arXiv:2610.10317v1 Announce Type: new Abstract: Tabular foundation models (TFMs) are pretrained across diverse tabular tasks and make predictions on a new table at inference time using its labeled exa

researcharxiv-cs-lg

8 Oct 2026

Transferability and operational reliability of a Prithvi crop classification foundation model under phenological and geographic shift across three continents

arXiv:2610.08810v1 Announce Type: new Abstract: Fine-tuned geospatial foundation models (GeoFMs) pretrained on large satellite archives have been shown to improve crop classification accuracy and geog

8 Oct 2026

Adversarially Trained Linear Transformers Are Optimal Robust In-Context Learners for Gaussian Mixtures

arXiv:2610.07754v1 Announce Type: cross Abstract: Adversarial training is one of the most reliable defenses against adversarial attacks, but its high computational cost must generally be paid anew for

Beyond Scalar IoU: Structured Verification from Rollout Groups for Video Temporal Grounding

arXiv:2610.07601v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) provides a natural framework for adapting pretrained models to video temporal grounding, where gen

safetyarxiv-cs-ai

DensiTok: Making Feed-Forward 3D Gaussian Splatting See More Views Than It Is Given

arXiv:2610.07958v1 Announce Type: new Abstract: Feed-forward 3D Gaussian Splatting (3DGS) reconstructs a scene in a single forward pass, replacing per-scene optimization with a network trained across

Forensic Reserve: Eliciting Latent Knowledge for Image Forgery Detection

arXiv:2610.08639v1 Announce Type: new Abstract: As generated images become increasingly realistic, reliable forgery detection is essential for maintaining trust in visual information. However, existin

From the Drosophila Visual Connectome to General-Purpose Computer Vision

arXiv:2610.08418v1 Announce Type: new Abstract: Biological connectomes encode structured solutions to visual computation that may provide reusable inductive biases for artificial vision. We develop Co

GeneICL: A Tabular Foundation Model for Bulk Transcriptomics

arXiv:2610.08694v1 Announce Type: new Abstract: Gene expression is widely measured in biomedicine, yet clinical outcome prediction remains challenging due to high dimensionality, strong feature correl

Investigating Model Compression for Neural Machine Translation in the Biomedical Domain

arXiv:2610.07032v1 Announce Type: cross Abstract: Large-scale pretrained transformer models have achieved state-of-the-art performance across diverse machine translation tasks, including multilingual

researcharxiv-cs-ai

JIVEAdapter: A Multi-Task Additive Low-Rank Adapter via Joint and Individual Variation Explained (JIVE)

model-releasesarxiv-cs-ai

arXiv:2610.07036v1 Announce Type: new Abstract: Parameter-efficient fine-tuning adapts pretrained models at a fraction of the cost of full fine-tuning, yet most low-rank adapters are single-task and r

SpecBraM: What Should an EEG Foundation Model Predict? Masked Band-Power Prediction versus Waveform Reconstruction

arXiv:2610.07484v1 Announce Type: new Abstract: Self-supervised EEG models often reconstruct masked waveforms or predict discrete codes. We study a task-aligned alternative: masked band-power predicti

Towards In-Parameter Memory Augmentation for Large Language Models

model-releasesarxiv-cs-cl

arXiv:2610.08630v1 Announce Type: new Abstract: Recently Large Language Models (LLMs) and LLM-based agents increasingly need to incorporate knowledge acquired after pretraining, e.g., domain facts, us

WildMatch: Weakly Supervised Image Matcher Adaptation for Wildlife Re-Identification

arXiv:2610.07384v1 Announce Type: cross Abstract: Individual animal re-identification from camera-trap imagery is an instance retrieval problem central to non-invasive wildlife monitoring: a query ima

tutorialsarxiv-cs-ai

FACET: Factorized Asymmetric Conditioning for Efficient Transport in High-Fidelity Fluorescence Microscopy Synthesis

arXiv:2610.05353v1 Announce Type: cross Abstract: Fluorescence microscopy reveals where proteins localize, but only a limited number of proteins can be imaged in the same cell; generating these images

Frequency-Decoupled Diffusion Guidance for Non-Blind Image Deblurring

arXiv:2610.06221v1 Announce Type: new Abstract: Pretrained diffusion models provide powerful image priors for training-free posterior sampling in image restoration. To guide this sampling process, fre

tutorialsarxiv-cs-cv

Kandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio Generation

arXiv:2610.05608v1 Announce Type: new Abstract: We present Kandinsky 6.0 Video, a family of foundation diffusion models for synchronized text-to-audio-video generation, comprising Kandinsky 6.0 Video

safetyarxiv-cs-cv

Learning Discriminative Geometry for Drifting Models

arXiv:2610.04703v1 Announce Type: cross Abstract: Recently proposed Drifting Models shift iterative distribution refinement from inference to training, enabling effective one-step generation. However,

Learning to Learn a Language

arXiv:2610.05879v1 Announce Type: cross Abstract: We present the Prior-Fitted Language Model (PFLM), a 300M-parameter byte-level transformer pretrained only on samples from a synthetic non-linguistic

Local Ai

Localize-and-Detect: Auditing Task-Level Poisoning in Instruction-Tuned Models

arXiv:2610.03960v1 Announce Type: cross Abstract: Instruction fine-tuning adapts a pretrained language model to follow instructions by training it on instruction--response pairs from a collection of t

local-aiarxiv-cs-lg

ReMaD: Tuning-free Domain Adaptation for Classification and Out-of-Distribution Detection

arXiv:2610.05718v1 Announce Type: new Abstract: We introduce Reduced-rank Mahalanobis Distance (ReMaD), a novel prototypical distance-based refinement to classification and out-of-distribution (OOD) d

Render to Reason: Novel-View Semantic Prediction Improves Spatial Understanding in VLMs

arXiv:2610.05417v1 Announce Type: new Abstract: Recent works augment Vision-Language Models with geometry features from pretrained 3D models, expecting that the geometric signal will boost spatial rea

Local Ai

Robust Ensemble Guidance for Scientific Inverse Problems

arXiv:2610.05371v1 Announce Type: new Abstract: Ensemble guidance combines pretrained diffusion priors with black-box forward models to solve inverse problems without differentiating through the physi

local-aiarxiv-cs-lg

Separators Make Carry Propagation Learnable:The Geometry of Latent Carry in a Multiplication Transformer

arXiv:2610.06605v1 Announce Type: new Abstract: Transformers asked to multiply multi-digit numbers in a single forward pass often fail, and interpretability studies of pretrained language models find

Task Vector Descent: Learning from Non-IID Batches

arXiv:2610.05402v1 Announce Type: new Abstract: A central challenge in continual learning is to acquire new knowledge without forgetting what the model has already learned. This challenge appears in l

To Learn is to Wander: Learning Across Graphs and Tasks with Random Walks

arXiv:2610.06694v1 Announce Type: new Abstract: Graph foundation models aim to transfer across graphs, feature spaces, relational schemas, and prediction tasks, yet existing approaches typically gener

tutorialsarxiv-cs-lg

Hardware

WAMJET: A Harness for World Action Model Acceleration

arXiv:2610.03797v1 Announce Type: new Abstract: World Action Models (WAMs) leverage pretrained video foundation models for robot manipulation, but their large backbones and video-action co-prediction

hardwarearxiv-cs-cv

WASP: Weakly Aligned Spatiotemporal Pairs for Fetal Brain MRI-Ultrasound Learning

arXiv:2610.04601v1 Announce Type: new Abstract: Magnetic Resonance Imaging (MRI) is widely regarded as the optimal sensor for fetal brain analysis due to its superior soft-tissue contrast and anatomic

safetyarxiv-cs-cv

AREX: Affine-Residual Exponential Integrator for Few-Step Sampling in Flow Matching

arXiv:2610.03483v1 Announce Type: cross Abstract: We introduce AREX, a training-free sampler for pretrained flow matching models that uses the target mean and covariance to capture an analytically tra

researcharxiv-cs-ai

Bounded Reachability & Jailbreak Detection via Contraction-Constrained State Space Models

arXiv:2610.02853v1 Announce Type: new Abstract: Safety heads are lightweight classifiers attached to pretrained language models for flagging harmful inputs before generation. Their empirical detection

safetyarxiv-cs-ai

ENCORE: Exact Non-equilibrium COntrol with Replica Exchange for Diffusion Generation

arXiv:2610.02538v1 Announce Type: cross Abstract: Inference-time control steers a pretrained generative model towards a target distribution without retraining. We study tilted targets pi_0propto G_0,p

Geometry-Aware Time Reparameterization for Flow-Map Distillation

arXiv:2610.02427v1 Announce Type: cross Abstract: Flow-map distillation enables one- and few-step generation by learning finite-time transitions of a pretrained generative ODE. We investigate whether

tutorialsarxiv-cs-ai

GRAFT: Growing Agglomerative Foundation Models via Continual Teacher Distillation

arXiv:2610.02597v1 Announce Type: new Abstract: Vision foundation models such as DINOv2, SigLIP2, and MASt3R develop complementary capabilities from different pretraining objectives, yet their knowled

Self-Repairing Recurrent Ensembles for Real-Time Recovery from Distribution Shift

model-releasesarxiv-cs-ro

arXiv:2610.03249v1 Announce Type: cross Abstract: Deploying a pretrained controller exposes it to conditions that are absent from its training data. Sensor drift, outright sensor failure and accumulat

XGenAct: Geometry-Enhanced World Action Models through Cross-Task Generation

arXiv:2610.03516v1 Announce Type: cross Abstract: World action models (WAMs) have advanced robot control by predicting how observations and actions evolve over time. Despite this progress, RGB and act

tutorialsarxiv-cs-cv

Beyond Decodability: Do Acoustic Factors Drive Predictions in Speech-Based Alzheimer's Assessment?

arXiv:2610.01846v1 Announce Type: cross Abstract: Speech-based Alzheimer's disease (AD) assessments increasingly rely on pretrained self-supervised learning (SSL) models that learn acoustic representa

safetyarxiv-cs-lg

Beyond Domain-Level Adaptation: Margin-Oriented Semantic-Appearance Interaction Correction for Personalized Federated Vision-Language Models

arXiv:2610.01625v1 Announce Type: new Abstract: Federated parameter-efficient fine-tuning enables distributed clients to adapt pretrained vision-language models without sharing raw data or updating th

Closing the Loop: Practical Training Recipes for Looped Language Models

model-releasesarxiv-cs-ai

arXiv:2610.00673v1 Announce Type: cross Abstract: Looped language models increase effective depth by repeatedly applying a shared block of layers, but existing large-scale recipes require multi-stage

Continual Concept Erasure in Diffusion Models by Suppressing Cross-Edit Interference

arXiv:2610.01989v1 Announce Type: new Abstract: Concept erasure removes copyright-protected, privacy-sensitive, or otherwise undesirable concepts from pretrained text-to-image diffusion models to supp

Coupling Perception and Reasoning in Federated Multimodal Graph Foundation Models

arXiv:2610.00277v1 Announce Type: new Abstract: Federated multimodal graph foundation models (GFMs) aim to adapt pretrained multimodal models to decentralized graph data, where each client owns a priv

Faynt: Scaling and Optimizing Policies for Competitive Melee

arXiv:2610.02144v1 Announce Type: new Abstract: We introduce Faynt, a family of 10M- and 75M-parameter Transformer policies for Super Smash Bros. Melee, each controlling all 26 characters with a singl

IrekoGPT: Turning Structured Pruning into Post-Hoc Slimmable LLMs

arXiv:2610.00426v1 Announce Type: new Abstract: We introduce IrekoGPT, a post-hoc method for converting pretrained LLMs into slimmable models whose width can be adjusted at inference time. Building on

JEPA-TTT: Persistent Test-Time Training of Latent World Models for Planning under Dynamics Shifts

arXiv:2610.00722v1 Announce Type: new Abstract: World models enable agents to plan by predicting future states of the environment, but their predictions can become unreliable when test-time dynamics d

Learning Transferable Skills using Goal-Conditioned Bisimulation

arXiv:2610.00676v1 Announce Type: cross Abstract: Unsupervised skill discovery has emerged as a promising approach for leveraging reward-free datasets to pretrain general-purpose policies. However, cu

tutorialsarxiv-cs-ai

Low-Budget Active Learning through Entropic Optimal Transport

arXiv:2610.01199v1 Announce Type: new Abstract: We consider low-budget active learning, which consists of selecting a limited number of points, the coreset, such that a model can be trained to high ac

RASteer: Retain-Aware Activation Steering for Concept Erasure in Diffusion Models

arXiv:2610.01969v1 Announce Type: new Abstract: Concept erasure aims to remove a target concept, such as a copyrighted style, a recognizable character, or unsafe content, from a pretrained text-to-ima

Specificity-Aware Diffusion Steering via Variance-Reduced Sequential Monte Carlo

arXiv:2610.00395v1 Announce Type: new Abstract: Inference-time steering enables pretrained diffusion models to satisfy new constraints without full retraining. However, specificity-aware generation is

STEER: Reducing Inference Cost in Relational Foundation Models through Semantically Informed Sampling

arXiv:2610.00907v1 Announce Type: cross Abstract: Relational foundation models (RFMs) are pretrained once on a collection of relational databases and prediction tasks, and then applied zero-shot to pr