探索全球前沿学术脉络

Doppler-enhanced superheterodyne Rydberg microwave receiver

arXiv:2606.24247v1 Announce Type: cross Abstract: We report the enhanced sensitivity of the Rydberg microwave (MW) receiver by exploiting the Doppler effect in a vapor cell. A two-photon Rydberg ladder scheme is implemented via the co-propagation of probe and coupling lasers, which enhances the Doppler effect. When an MW field is applied, microwave dressing modifies the velocity-dependent resonance condition, enabling stronger contributions from atoms with non-zero velocities and leading to an enhancement of the EIT transmission. Based on this mechanism, we achieve a sensitivity of $35.1\ \mathrm{nV\ cm^{-1}\ Hz^{-1/2}}$ using the heterodyne technique, which is 1.5 times better than that obtained in the counter-propagating configuration. Meanwhile, the required local oscillator (LO) field is reduced by a factor of 17.6 compared with the counter-propagating configuration, which is advantageous for applications requiring minimal radiation and low power consumption. Moreover, the co-propagating configuration is more amenable to integration or portable sensing platforms because multiple laser fields can be delivered through a single optical fiber.

PACT: Privileged Trace Co-Training for Multi-Turn Tool-Use Agents

Multi-turn tool-use agents must reason, call tools, and adapt to observations across several interaction turns. Post-training such agents is challenging, as reinforcement learning often suffers from sparse rewards and weak credit assignment despite matching the prompt-only inference setting, while supervised fine-tuning on expert traces provides dense process supervision but can over-constrain the model to fixed trajectories. To tackle this, we propose PACT, a Privileged trAce Co-Training framework for multi-turn tool-use agents. The key idea is to use expert traces only as training-time optimization signals rather than rollout-time hints. PACT keeps rollout generation prompt-only, then uses expert traces to guide optimization through two complementary signals: a trace-conditioned RL surrogate that evaluates prompt-only rollouts under expert-trace context, and a component-aware SFT loss that supervises reasoning prefixes and tool-calls with annealed strength. To reduce over-reliance on the training-only trace context, PACT further introduces a prompt-only anchoring. We also provide a latent-trace view that connects the two trace-based objectives and explains how expert traces can guide optimization without being used during rollout generation. Experiments on FTRL, BFCL, and ToolHop show that PACT consistently improves over strong SFT- and RL-based baselines, highlighting the value of privileged trace co-training for multi-turn tool-use learning.

Playful Agentic Robot Learning

arXiv:2606.19419v1 Announce Type: cross Abstract: Current agentic robot systems can write executable Code-as-Policy programs, observe feedback, and revise behavior across multiple attempts, but they remain largely task-driven: reusable skills are acquired only after explicit instructions. We study Playful Agentic Robot Learning, where an embodied coding agent uses self-directed play as a continual skill-learning stage before downstream tasks arrive. We introduce RATs, Robotics Agent Teams designed for play-time skill acquisition. During play, RATs proposes novel yet learnable exploratory tasks, plans and executes robot-code policies, verifies intermediate progress, diagnoses failures, retries with dense, step-level feedback, and distills successful executions into a persistent code skill library. At test time, the agent reuses relevant skills from this frozen library to help solve new tasks. Experiments in LIBERO-PRO and MolmoSpaces show that play-learned skills improve held-out downstream tasks over no-play and random-play baselines, with 20.6 and 17.0 percentage-point gains over CaP-Agent0 on LIBERO-PRO and MolmoSpaces, respectively. Moreover, the learned skills can be plugged into other inference-time Code-as-Policy agents by simply retrieving them into the context, improving RoboSuite and real-world transfer by 8.9 and 8.8 points, respectively, without finetuning the underlying model.