Large language models (LLMs) have demonstrated impressive performance on several tasks and are increasingly deployed in real-world applications. However, especially in high-stakes settings, it becomes vital to know when the output of an LLM may be unreliable. Depending on whether an answer is trustworthy, a system can then choose to route the question to another expert, or otherwise fall back on a safe default behavior. In this work, we study the extent to which LLMs can reliably indicate confidence in their answers, and how this notion of confidence can translate into downstream accuracy gains. We propose Self-Reflection with Error-based Feedback (Self-REF), a lightweight training strategy to teach LLMs to express confidence in whether their answers are correct in a reliable manner. Self-REF introduces confidence tokens into the LLM, from which a confidence score can be extracted. Compared to conventional approaches such as verbalizing confidence and examining token probabilities, we demonstrate empirically that confidence tokens show significant improvements in downstream routing and rejection learning tasks.

In standard reinforcement learning setups, the agent receives observations and performs actions at evenly spaced intervals. However, in many real-world settings, observations are expensive, forcing agents to commit to courses of action for designated periods of time. Consider that doctors, after each visit, typically set not only a treatment plan but also a follow-up date at which that plan might be revised. In this work, we formalize the setup of timing-as-an-action. Through theoretical analysis in the tabular setting, we show that while the choice of delay intervals could be naively folded in as part of a composite action, these actions have a special structure and handling them intelligently yields statistical advantages. Taking a model-based perspective, these gains owe to the fact that delay actions do not add any parameters to the underlying model. For model estimation, we provide provable sample-efficiency improvements, and our experiments demonstrate empirical improvements in both healthcare simulators and classical reinforcement learning environments.

Recent work in Human-Robot Interaction (HRI) has shown that robots can leverage implicit communicative signals from users to understand how they are being perceived during interactions. For example, these signals can be gaze patterns, facial expressions, or body motions that reflect internal human states. To facilitate future research in this direction, we contribute the \textttREACT database, a collection of two datasets of human-robot interactions that display users' natural reactions to robots during a collaborative game and a photography scenario. Further, we analyze the datasets to show that interaction history is an important factor that can influence human reactions to robots. As a result, we believe that future models for interpreting implicit feedback in HRI should explicitly account for this history. \textttREACT opens up doors to this possibility in the future.

Rates of missing data often depend on record-keeping policies and thus may change across times and locations, even when the underlying features are comparatively stable. In this paper, we introduce the problem of Domain Adaptation under Missingness Shift (DAMS). Here, (labeled) source data and (unlabeled) target data would be exchangeable but for different missing data mechanisms. We show that if missing data indicators are available, DAMS reduces to covariate shift. Addressing cases where such indicators are absent, we establish the following theoretical results for underreporting completely at random: (i) covariate shift is violated (adaptation is required); (ii) the optimal linear source predictor can perform arbitrarily worse on the target domain than always predicting the mean; (iii) the optimal target predictor can be identified, even when the missingness rates themselves are not; and (iv) for linear models, a simple analytic adjustment yields consistent estimates of the optimal target parameters. In experiments on synthetic and semi-synthetic data, we demonstrate the promise of our methods when assumptions hold. Finally, we discuss a rich family of future extensions.

Humans expect robots to learn from their feedback and adapt to their preferences. However, there are limitations with how humans provide feedback to robots, e.g., humans may give less feedback as interactions progress. Therefore, it would be advantageous if robots could influence humans to provide more feedback during interactions. We conducted a 2x2 between-subjects user study (N=71) to investigate whether the framing and timing of a robot's reminder to provide feedback could influence human interactants. Human-robot interactions took place in the context of Space Invaders, a fast-paced and continuous collaborative environment. Our results suggest that reminders can influence the amount of feedback humans provide to robots, how participants feel about the robot, and how they feel about providing feedback during the interaction.

Due to the inefficiency of multiple binary images encryption, a parallel binary image encryption framework based on the typical variants of spiking neural networks, spiking neural P (SNP) systems is proposed in this paper. More specifically, the two basic units in the proposed image cryptosystem, the permutation unit and the diffusion unit, are designed through SNP systems with multiple channels and polarizations (SNP-MCP systems), and SNP systems with astrocyte-like control (SNP-ALC systems), respectively. Different from the serial computing of the traditional image permutation/diffusion unit, SNP-MCP-based permutation/SNP-ALC-based diffusion unit can realize parallel computing through the parallel use of rules inside the neurons. Theoretical analysis results confirm the high efficiency of the binary image proposed cryptosystem. Security analysis experiments demonstrate the security of the proposed cryptosystem.

In several medical decision-making problems, such as antibiotic prescription, laboratory testing can provide precise indications for how a patient will respond to different treatment options. This enables us to "fully observe" all potential treatment outcomes, but while present in historical data, these results are infeasible to produce in real-time at the point of the initial treatment decision. Moreover, treatment policies in these settings often need to trade off between multiple competing objectives, such as effectiveness of treatment and harmful side effects. We present, compare, and evaluate three approaches for learning individualized treatment policies in this setting: First, we consider two indirect approaches, which use predictive models of treatment response to construct policies optimal for different trade-offs between objectives. Second, we consider a direct approach that constructs such a set of policies without intermediate models of outcomes. Using a medical dataset of Urinary Tract Infection (UTI) patients, we show that all approaches learn policies that achieve strictly better performance on all outcomes than clinicians, while also trading off between different objectives. We demonstrate additional benefits of the direct approach, including flexibly incorporating other goals such as deferral to physicians on simple cases.

The Rensselaer Mandarin Project enables a group of foreign language students to improve functional understanding, pronunciation and vocabulary in Mandarin Chinese through authentic speaking situations in a virtual visit to China. Students use speech, gestures, and combinations thereof to navigate an immersive, mixed reality, stylized realism game experience through interaction with AI agents, immersive technologies, and game mechanics. The environment was developed in a black box theater equipped with a human-scale 360◦ panoramic screen (140h, 200r), arrays of markerless motion tracking sensors, and speakers for spatial audio.

In this paper, we propose the multichannel variable-radius multiple access (M-VRMA) scheme for power-controlled multichannel medium access control (MAC) in mobile ad hoc networks. We propose the RTS/object-to-sending (OTS)/VP-CTS (ROV) protocol based on M-VRMA, multiple access with lag time, OTS, and VP-CTS for efficient variable-radius supports in heterogeneous ad-hoc networks, without relying on busy-tone or any mechanisms that require expensive specialized hardware or more than one transceivers per device. Our simulation results demonstrate that for the same radio bandwidth, ROV can achieve considerably higher throughput as compared to fixed-radius IEEE 802.11 or previous power-controlled RTS/CTS protocols without variable-radius supports.

We propose time division multiple access with circular reservation (TDMA/CR), a time/space division multiple access (TSDMA)-based medium access control (MAC) protocol for wireless mobile networks with control units, including wireless LANs, cellular networks with ad hoc relaying capability, and ad hoc networks with access points or clusterheads. Different from the MAC protocol of IEEE 802.11, TDMA/CR is centralized and can utilize the computation capability of base stations or access points to increase network throughput, reduce latency, and provide QoS guarantees. We evaluate the performance of TDMA/CR and show that the utilization achievable by the wireless-tree (or splitting) channel access mechanisms of TDMA/CR is about 40% to 46% and the channel access delay is small and bounded. Moreover, TDMA/CR can achieve considerably higher throughput (e.g., by a factor of about 3 or higher) due to its support for variable-radius transmissions in ad hoc wireless networks.