Found 11 document(s)
Page 1 of 1
-
Rubric-based reinforcement learning (RL) trains policies for open-ended generation with natural-language criteria as rewards, but these rewards can become ineffective as training progresses. We identify two failure modes...
-
Practical image generation requires both human preference alignment and few-step efficiency, yet combining reinforcement learning (RL) with distillation is often multi-stage and may erode RL gains We observe that few-ste...
-
Cinematic multi-shot video generation requires both narrative planning and iterative refinement. Existing methods either lack narrative reasoning or rely on feed-forward workflows without policy updates. In this paper, w...
-
Vision-language-action (VLA) models have shown promise in robotic manipulation, but imitation learning (IL) alone often struggles to master complex long-horizon skills. Reinforcement learning (RL) offers an effective ave...
-
Existing RGBA generation methods commonly learn a transparency-specific representation through reconstruction, introducing an additional training stage and potentially altering the RGB decoding function inherited from pr...
-
China’s export performance is characterized by two striking features: a steady rise in export sophistication and the continued prominence of less sophisticated products. This paper proposes the “Many Chinas” hypothesis, ...
-
In switchable spin crossover (SCO) molecular materials, control of their critical temperature and cooperativeness is essential for practical applications. This requires systematic fine-tuning of the molecular packing in ...
-
A determination of the jet energy scale is presented using proton$-$proton collision data with a centre-of-mass energy of $\sqrt{s}=13$ TeV, corresponding to an integrated luminosity of 140 fb$^{-1}$ collected using the ...
-
To examine the origins, characteristics, and outcomes of resilient governance with Chinese characteristics, this open access book takes Guangzhou, a typical Chinese city from 1990 to 2015, as an example. Through particip...
-
We use acoustic resonances in a planar structure to convert orbital angular momentum (OAM) into an axisymmetric sound field. The resonances twist wave vector to produce the OAM needed spiral phase. We numerically and the...