…Rohan Surana , , , , , , , , , , , Abstract A unified framework combines candidate generation and ranking in a single autoregressive model using factorized group-relative policy optimization to address credit assignment challenges in end-to-end retrieval…
…Adaptive Failure-Informed Learning for Vision-Language-Action Models (2026) Pre-VLA: Preemptive Runtime Verification for Reliable Vision-Language-Action and World-Model Rollouts (2026) Dynamic Execution Commitment of Vision-Language-Action…
…We convert a bidirectional video generator into a few-step autoregressive world model using causal forcing , DMD-style distillation , and long-rollout training . Training on self-generated long-horizon contexts exposes the…
…large language models (LLMs) is entering a new era. Rather than relying on a static recipe, EvoTrainer enables LLM policies and their training harnesses to evolve jointly over time. This is more…
…Self-Corrective Autoregressive World Model for Long-horizon Driving Simulation (2026) RAVEN: Real-time Autoregressive Video Extrapolation with Consistency-model GRPO (2026) Delta Forcing: Trust Region Steering for Interactive Autoregressive Video Generation…
…Aligning Diffusion Models with Fine-Tuning Is Easier Than You Think (2026) Reward Sharpness-Aware Fine-Tuning for Diffusion Models (2026) Step-level Denoising-time Diffusion Alignment with Multiple Objectives (2026) Astrolabe…
…Real-time Autoregressive Video Extrapolation with Consistency-model GRPO (2026) StreamChar: Long-Horizon Streaming Character Audio-Video Generation with Decoupled Orchestration (2026) Please give a thumbs up to this comment if you…
…Hidir Yesiltepe , , , , , , Abstract VideoMLA reduces memory usage in video diffusion models by replacing per-head keys and values with shared low-rank content and decoupled 3D-RoPE positional keys, maintaining quality while…
…multi-agent rollout through shared visual prefix and KV cache reuse . Across V*, CountBench, the RefCOCO family, and HallusionBench, Visual Para-Thinker++ consistently outperforms single-trajectory and inference-time parallel baselines, with…
…In GPT-style language-model pretraining , PACI matches the stability and final perplexity of synchronous 1F1B-flush , retains the same peak memory footprint , achieves fully utilized pipeline throughput, and improves training time…