LoRA, PEFT, instruction tuning and domain adaptation — adapting a pretrained model without paying to train one.
arXiv:2610.09756v1 Announce Type: new Abstract: Classical Arabic poetry generation requires simultaneously satisfying semantic, linguistic, and fine-grained prosodic constraints. Existing systems typ...
arXiv:2610.08818v1 Announce Type: new Abstract: Spatio-temporal forecasting is a cornerstone of logistics, urban planning, and intelligent transportation systems. However, constrained by deployment c...
arXiv:2610.10429v1 Announce Type: new Abstract: Autoregressive video generation requires denoising the current frames while writing their key-value representations as context for future predictions....
Latent reasoning models repeatedly update a latent state using the same recurrent block. As the recurrent depth increases, the sequence of latent states may converge to a compact subset of the state s...
Text-to-image diffusion models are increasingly distilled into few-step variants and being deployed to enable fast inference. However, their ability to generate harmful or undesired content poses sign...
Prompted embedding models have recently received increasing attention, particularly for retrieval, where detailed retrieval instructions are provided as part of the retrieval prompt. Several new datas...
Terminal-agent capability depends jointly on model weights and the runtime harness that formats prompts, binds tools, and handles error recovery. Existing harness-model co-evolution approaches improve...
Instruction-tuned models are deployed into environments where domains are heterogeneous and evolve, yet adding new domains or data typically requires costly retraining. We present AutoAdapt, a modular...
Large Language Models (LLMs) inevitably internalize substantial amounts of sensitive or private information during pre-training, while LLM unlearning aims to selectively erase specific knowledge to pr...
Fine-tuning-as-a-service enables users to adapt aligned large language models (LLMs) to specialized tasks, but malicious fine-tuning can erode refusal behavior while preserving task performance on leg...
Pretrained models for cell and nuclear instance segmentation differ substantially in architecture, pretraining data and objectives, parameter count, inference strategy, adaptation requirements, postpr...
Machine-learned interatomic potentials (MLIPs) excel at in-distribution tasks, accelerating drug and material development, yet they struggle to generalize out-of-distribution. We propose to push the c...
Model organisms of alignment-relevant behaviors (e.g., backdoors, sycophancy, spurious correlations) have emerged as a key tool for evaluating whitebox interpretability techniques. We argue that the p...
When does distribution alignment help a frozen foundation-model embedding generalize across acoustic domains? For cross-domain mosquito-species classification we report a strength-monotonic law: the s...
Spoken language identification (LID) aims to recognize the target language regardless of accent. In practice, however, LID models fine-tuned from self-supervised speech representations frequently conf...
The paper introduces OFAG, a foundation model designed for attributed graph clustering that can be trained once and applied to diverse graphs without graph‑specific tuning. It learns a reusable clustering strategy from synthetic graphs and uses a dimension‑agnostic encoder to handle varying feature spaces, producing clustering‑friendly node representations in a single forward pass. Across ten datasets, a frozen OFAG model outperforms baselines in both speed and clustering quality, and the authors provide code and pretrained checkpoints for easy adoption.
arXiv:2610. 07529v1 Announce Type: new Abstract: Wave-based processors promise fast, energy-efficient Fourier layers for neural operators.
QF3 (Fast Flow RL with Filtered Q-Gradients) is an online off‑policy reinforcement learning algorithm that trains flow policies by combining flow matching with the critic’s action gradient, backpropagated through a one‑step prediction of the flow output. It selectively applies the critic gradient only to action dimensions close to the replay action to ensure reliable updates. QF3 is the first off‑policy flow RL method to train humanoid locomotion policies from scratch and transfer them zero‑shot to hardware, achieving a 10× wall‑clock speedup over the on‑policy FPO++ method, and it can also fine‑tune pretrained flow‑based manipulation policies on ABC‑Sim and Robomimic tasks.
The paper introduces Contrastive‑SDXL, a day‑to‑night augmentation framework that uses SDXL‑Turbo and Low‑Rank Adaptation to generate realistic night‑time images for pedestrian detection. It employs a patch‑wise semantic contrastive loss guided by a pretrained DINOv2 encoder and an object consistency loss to preserve pedestrian and local semantic structure during translation, achieving a FID of 22.5. Detectors trained on these synthetic images show a 6‑7% reduction in miss rate compared to a daytime‑only baseline, approaching the performance of models trained on real night‑time data.