Skip to content

Author

Seonghyun Jin

7 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Preprint Oct 2026

FastOPD: On-Policy Distillation for Lightweight VLA Deployment

Vision-Language-Action (VLA) foundation models have scaled rapidly to enhance manipulation performance and generalizability, but this scaling incurs high computational costs that render real-world deployment increasingly challenging. Existing approaches typically mitigate this issue by designing smaller architectures o...

Yoojin Oh, Jeongsol Kim, Yeonwoo Seo et al. · 0 citations
#artificial intelligence Preprint Sep 2026

DriftOPD: Sequence-Level Reverse-KL Distillation for One-Step VLA Policies

Vision-Language-Action (VLA) models increasingly rely on action experts that generate short action chunks under receding-horizon control. While chunk-level training is convenient across robot embodiments, it optimizes local action likelihood without explicitly accounting for long-horizon task success. Sequence-level re...

Youngjun Jun, Kyumin Choi, Young Min Kim et al. · 0 citations
#machine learning Preprint Sep 2026

Adjoint Guidance Flow: Amortized Critic Guidance for VLA Policies

Adjoint Guidance Flow is proposed, which is a deterministic optimal control problem, whose optimal guidance is a costate that carries the terminal critic gradient back through the remaining flow, and regress the guidance network onto this costate while keeping both the VLA and critic frozen.

Jeongsol Kim, Youngjun Jun, Kyumin Choi et al. · 0 citations
#artificial intelligence Preprint Sep 2026

Don't Throw Away the Tail: Action Upcycling for Policy Acceleration

This work proposes Action Upcycling, a training-free algorithm that reuses actions the policy would otherwise discard, without accessing model internals or drawing extra samples, and finds that discarded actions stay close to their replanned versions as long as the action velocity remains smooth.

Taesung Kwon, Jangho Park, Sunwoo Park et al. · 0 citations

Dynamic Manipulation with World-Action Models via Counterfactual Planning

DPP enables real-time dynamic manipulation on a single consumer GPU without additional training on dynamic data and constructs a counterfactual observation that places a predicted target position in a familiar robot context, allowing the model to invoke an existing manipulation skill rather than generate a recovery beh...

Sunwoo Park, Won-Sang Lee, Seonghyun Jin et al. · 1 citation

CRePE: Curved Ray Expectation Positional Encoding for Unified-Camera-Controlled Video Generation

Curved Ray Expectation Positional Encoding (CRePE) is introduced, which represents each image token as a depth-aware distribution along its Unified Camera Model (UCM) ray and integrates the expected rotary positional phasor along the curved path this distribution traces when projected into each query view.

Seonghyun Jin, Young Min Kim, Sunwoo Park et al. · 3 citations

FILT3R: Latent State Adaptive Kalman Filter for Streaming 3D Reconstruction

FILT3R, a training-free latent filtering layer that casts recurrent state updates as stochastic state estimation in token space, is proposed, and it is demonstrated that gains shrink in stable regimes as uncertainty contracts with accumulated evidence, and rise when genuine scene change increases process uncertainty.

Seonghyun Jin, Jong Chul Ye · 3 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.