Engineering
Video Generative Models as Geometry Learner
GeoNeXt repurposes pretrained video generative models for geometry-estimation tasks such as depth and surface-normal prediction by formulating geometry estimation as…
Engineering
PonderPounce: A Pretrained MLLM as an Episode Context Engine for Robot Control
PonderPounce is a robot-control system using a pretrained multimodal LLM as episodic memory: “Ponder,” a reasoning-focused MLLM, accumulates observations and…
Engineering
Rubric-to-Code Credit Assignment for Reinforcement Learning
Rubric-to-Code Credit Assignment (RCCA) is a reinforcement-learning framework that converts functional feedback from rubrics into localized optimization signals for generated…
Engineering
Language Chain in Alignment: Cross-lingual Ranking Preference Optimization
Cross-lingual Ranking Preference Optimization (CRPO) transfers preference knowledge from English to improve LLM alignment in other languages, using a hierarchical…
Engineering
Acquire, Repair, Preserve: A Diagnosis-Guided Post-Training Recipe for Small-Model Dialogue Game Agents
“Acquire, Repair, Preserve” is a three-stage post-training recipe for small language models playing interactive dialogue games: supervised fine-tuning for broad…
Engineering
GGSS: Geodesic-Gated Spherical Steering for Inference-Time Debiasing of Generative Vision-Language Models
GGSS (Geodesic-Gated Spherical Steering) is an inference-time debiasing method for generative vision-language models that discovers a counterfactual bias subspace on…
Engineering
Ask or Answer: A Decision Framework for Multi-Turn Health Misinformation Intervention
Reward-Optimized Probe-and-Respond (RO-PnR) is a framework that decides when conversational agents should ask clarifying questions versus give direct corrections during…
Engineering
Training, learning and inference: unified dynamics of neural systems
This paper introduces a Generation-Fact Graph (GFG) framework unifying training, learning, and inference dynamics in neural systems, developing a predictive…
Engineering
Generative Semantic Scene Completion
Generative Semantic Scene Completion (GSSC) is a discrete-diffusion framework for outdoor LiDAR scene completion with three components: paired sparse-dense synthesis…
Engineering
Lost in Compression: A Controlled Cross-Lingual Audit of Extractive Prompt Compressors
This paper presents a cross-lingual evaluation of extractive prompt compressors across ten languages and eleven target models, benchmarking four learned…
Engineering
Sliding-window beats linear attention
This paper shows that Sliding Window Attention (SWA) with attention sinks delivers comparable or superior performance to post-trained Linear Attention…
Engineering
LMSM: LLM Security Framework Inspired by Linux Security Modules
Researchers propose Language Model Security Modules (LMSM), a framework that adapts the architecture of Linux Security Modules to protect large…