ForeTime-VLA: Causal Future-Token Distillation from a World Action Model for Conveyor-Belt Manipulation

• URL: http://arxiv.org/abs/2608.20735v2
• Date: Mon, 24 Aug 2026 03:31:07 GMT
• Title: ForeTime-VLA: Causal Future-Token Distillation from a World Action Model for Conveyor-Belt Manipulation
• Abstract summary: We introduce ForeTime-VLA, a dense pi0.5 policy that distills a future-aware, action-equivalent representation from a frozen Fast-WAM-derived teacher.<n>Online, an eight-frame history encoder predicts this target together with manipulation phase and normalized time-to-transition.<n>Test MAE decreases from 0.134119 to 0.130593 (2.63%); paired-bootstrap 95% CI: 0.82-4.48% improvement, and test L2 decreases by 3.02%, at a 2.46-2.93% latency cost.
• Score: 6.874305530359895
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Manipulating moving objects requires a policy to anticipate contact events, yet vision-language-action (VLA) policies are commonly fine-tuned from the current observation alone. World action models (WAMs) learn predictive dynamics, but running a video-scale teacher or explicitly imagining future frames at deployment is costly. We introduce ForeTime-VLA, a dense pi0.5 policy that distills a future-aware, action-equivalent representation from a frozen Fast-WAM-derived teacher while remaining causal at inference. Offline, current and future video latents are compressed into a whitened 64-D target. Online, an eight-frame history encoder predicts this target together with manipulation phase and normalized time-to-transition. Four future tokens and one phase token condition the VLM prefix, while the predicted future and transition horizon condition the action expert. Training retains the original flow-matching action target and adds cosine, relational geometry, phase, time-to-transition, and action-equivalence objectives. On a deduplicated conveyor-belt dataset, we compare 40k-step checkpoints on 768 matched windows per split. Test MAE decreases from 0.134119 to 0.130593 (2.63%; paired-bootstrap 95% CI: 0.82-4.48% improvement), and test L2 decreases by 3.02%, at a 2.46-2.93% latency cost. In quantitative real-robot evaluation, ForeTime-VLA achieves 81.1% stationary and 58.9% slow-moving grasp success, exceeding the next-best reference by 12.2 and 22.2 percentage points, respectively. Across three belt speeds, it completes 44/90 grasps versus 23/90 for pi0.5, including 11/30 versus 2/30 at fast speed. The agreement between offline orientation gains and reduced real-robot contact-pose failures supports causal future-token distillation as an effective way to improve dynamic manipulation without deploying the world-model teacher.

Related papers

• ReDraft, Don't Just Distill: Reference-Driven Revision for Continual VLLM Post-Training [66.72708893922605]
  Post-training of large multimodal models should add new capabilities while preserving those from pre-training.<n>SFT gives explicit target supervision that learns a task from near-zero accuracy, but its off-policy targets move the model far enough to cause forgetting.<n>We introduce ReDraft, which obtains both from the model's own failures.
  arXiv  Detail & Related papers  (2026-09-15T04:59:03Z)
• RynnValue: Scaling Robotic Value Foundation Models with Temporal Distance [84.26294415726723]
  General-purpose reward models are increasingly the bottleneck for scaling robot learning.<n>We introduce RynnValue, an open-source value foundation model for robotic manipulation.<n>RynnValue scales to over 7,000 hours and roughly 3M instruction-conditioned clips without preference or progress annotations.
  arXiv  Detail & Related papers  (2026-08-10T17:09:37Z)
• Discrete Action Space as a Prerequisite for GRPO Convergence in Small-Model Continuous Control [0.0]
  Group Relative Policy Optimization can fine-tune small language models for simulated quadrotor continuous-control tasks.<n>In our benchmark, vanilla GRPO fine-tuning of Qwen-0.5B for 25 Hz quadrotor velocity control collapses to the trivial zero action.<n>A high-fidelity simulation using Crazyflie 2.1 dynamics surfaces a hover-region training-distribution gap.
  arXiv  Detail & Related papers  (2026-07-07T17:14:19Z)
• Latent Clarity: Bridging World-Model Kinematics to Semantic Manifolds for Video Anomaly Anticipation [0.0]
  Continuous video detection is dominated by reactive Multiple Instance Learning (MIL) that collapses into features into scores.<n>We introduce PULS (Predictive Unified Latent Space), a continuous semantic world-model pipeline.<n>This translation alone yields a chunk-level AUROC of 0.8994 for UCF-Crime and 0.8162 for out-of-distribution- anomaly without MIL or hierarchical fusion.
  arXiv  Detail & Related papers  (2026-07-03T18:47:54Z)
• Intercepting the Future: Latent-Space Predictive World Model for Dynamic VLA Manipulation [6.790887409995267]
  AHEAD (Anticipatory Horizon Extrapolation with Adaptive Dynamics) is a predict-then-act wrapper that augments a frozen VLA with a motion-aware latent world model.<n>AHEAD adds 4.9M parameters to a frozen 7B OpenVLA and reaches 79 to 97% success across 20 dynamic simulation scenarios.
  arXiv  Detail & Related papers  (2026-06-01T16:55:38Z)
• ESPO: Early-Stopping Proximal Policy Optimization [78.79610718910628]
  ESPO (Early-Stopping Proximal Policy Optimization) detects trajectory failure on-the-fly and terminates rollouts early.<n>On DeepSeek-R1-Distill-Qwen-7B trained for mathematical reasoning, ESPO surpasses PPO on AIME2024 (46.28% vs. 45.25%), AMC2023 (85.83% vs. 82.94%), and MATH-500 (87.42% vs. 85.43%)
  arXiv  Detail & Related papers  (2026-05-28T12:40:22Z)
• CrossVLA: Cross-Paradigm Post-Training and Inference Optimization for Vision-Language-Action Models [7.489020109808801]
  We present CrossVLA, an empirical study of cross-paradigm Vision-Language-Action (VLA) post-training.<n>Three contributions: (i) a surrogate flow-matching log-probability estimator that lets DPO operate on continuous-action backbones without probability-flow ODE integration; (ii) a head-to-head comparison of LoRA and DoRA as the parameter-efficient layer for VLA DPO; and (iii) an inference-time anatomy showing the denoise loop dominates 78.6% of sample_actions latency and prefix-K/V caching a la
  arXiv  Detail & Related papers  (2026-05-21T01:02:41Z)
• Global Prior Meets Local Consistency: Dual-Memory Augmented Vision-Language-Action Model for Efficient Robotic Manipulation [95.89924101984566]
  We introduce OptimusVLA, a dual-memory VLA framework with Global Prior Memory (GPM) and Local Consistency Memory (LCM)<n>GPM replaces Gaussian noise with task-level priors retrieved from semantically similar trajectories.<n>LCM injects a learned consistency constraint that enforces temporal coherence and smoothness of trajectory.
  arXiv  Detail & Related papers  (2026-02-22T15:39:34Z)
• Time Is a Feature: Exploiting Temporal Dynamics in Diffusion Language Models [57.474294329887236]
  Diffusion large language models (dLLMs) generate text through iterative denoising.<n>Current decoding strategies discard rich intermediate predictions in favor of the final output.<n>We introduce two complementary methods that exploit temporal consistency.
  arXiv  Detail & Related papers  (2025-08-12T17:59:57Z)
• Spatiotemporal Contrastive Video Representation Learning [87.56145031149869]
  We present a self-supervised Contrastive Video Representation Learning (CVRL) method to learn visual representations from unlabeled videos. Our representations are learned using a contrasttemporalive loss, where two augmented clips from the same short video are pulled together in the embedding space. We study what makes for good data augmentations for video self-supervised learning and find that both spatial and temporal information are crucial.
  arXiv  Detail & Related papers  (2020-08-09T19:58:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.