| 1 |
Preserving Unstable Modes Through Inverse Dynamics in JEPA World Models |
通过逆动力学增强JEPA世界模型以保持不稳定模式 |
world model world models JEPA |
|
|
| 2 |
WASD: Wasserstein-based Knowledge Distillation for Large Language Models |
提出WASD以解决大语言模型蒸馏效率低下问题 |
distillation large language model instruction following |
✅ |
|
| 3 |
Independent Multi-Agent Reinforcement Learning with Counterfactual Semantic-Social World Models |
提出CASTLE框架以解决独立多智能体强化学习中的奖励信号模糊问题 |
reinforcement learning PPO world model |
|
|
| 4 |
MS-ECG-FM: Towards a More Universal Electrocardiogram Foundation Model for Health Monitoring using Multi-source Contrastive Learning |
提出MS-ECG-FM以解决ECG模型监督不足问题 |
representation learning contrastive learning foundation model |
|
|
| 5 |
Atom-JEPA: Joint-Embedding Predictive Architecture for 3D Atomistic Systems |
提出Atom-JEPA以解决3D原子系统自监督预训练问题 |
JEPA Joint-Embedding Predictive Architecture joint-embedding predictive architecture |
✅ |
|
| 6 |
Variance-Averse $n$-Step Offline Reinforcement Learning for Sparse Long-Horizon Environments |
提出VAN-Flow以解决稀疏长时间环境中的不可靠动作选择问题 |
reinforcement learning offline RL offline reinforcement learning |
|
|
| 7 |
Revisiting Temporal Regularization for Smooth Control in Deep Reinforcement Learning |
提出CATS方法以解决深度强化学习中的动作平滑性问题 |
reinforcement learning deep reinforcement learning |
|
|
| 8 |
TRACE: Rollout-Guided Quantization-Aware Training for FP4 Reinforcement Learning of MoE Language Models |
提出TRACE以解决FP4强化学习中的量化精度问题 |
reinforcement learning large language model |
|
|
| 9 |
Does On-Policy Distillation for Safety Pose Backdoor Risks? |
揭示OPD在安全性中的后门风险及其缓解方法 |
distillation large language model |
|
|
| 10 |
Optimization Encoders: Rethinking Second-Order Meta-Learning for Neural Fields |
提出优化编码器以改进神经场的元学习 |
representation learning latent optimization |
|
|
| 11 |
Reinforcement Learning with Conformal Action Sets: An Application to Sequential Recommendation |
提出RLCP以解决序列推荐中的固定动作集问题 |
reinforcement learning |
|
|
| 12 |
Reinforcement Learning for Hierarchical Reasoning Rewards: Minimax-Optimal Rates with Transformers |
提出层次化奖励的强化学习方法以优化推理任务 |
reinforcement learning |
|
|
| 13 |
FlowCF: Sparse Counterfactual Explanations for Mixed-Type Tabular Data using Flow Matching |
提出FlowCF以解决混合类型表格数据的稀疏反事实解释问题 |
flow matching |
|
|
| 14 |
Learning PDE solution operators with variable initial conditions via Latent Dynamics Networks |
提出变初始条件的潜在动力网络以解决PDE建模问题 |
latent dynamics |
|
|
| 15 |
Enhancing Diffusion Language Models with Autoregressive Post-Training Weights |
提出A2D框架以利用自回归后训练权重增强扩散语言模型 |
reinforcement learning instruction following |
|
|
| 16 |
On-Policy Distillation with Negative-Policy Rollouts |
提出负策略蒸馏以增强在线策略蒸馏的学习信号 |
distillation |
✅ |
|
| 17 |
Privileged Context as Drift in On-Policy Self-Distillation |
提出特权上下文作为漂移以优化在线自蒸馏训练 |
distillation |
|
|
| 18 |
Improving Synthetic Data Generation for Argument Mining via Adversarial Reinforcement Learning |
提出对抗强化学习框架以改善论证挖掘的合成数据生成 |
reinforcement learning |
|
|
| 19 |
AFA-BANDIT: Provably Near-Optimal Online Multi-Feature Classification Under Budget Constraints |
提出AFA-BANDIT以解决预算约束下的在线多特征分类问题 |
reinforcement learning deep reinforcement learning |
|
|