When Seekers Are Hard to Help: Evaluating Emotional Support Dialogue Systems in Worst-Case Interactions

Emotional Support Dialogue Systems (ESDSes) are increasingly evaluated and trained with LLM-simulated seekers. However, such simulated seekers often behave as cooperative, average-case users who disclose clearly, respond constructively, and accept support within a few turns. This can lead to overly optimistic evaluation and obscure whether ESDSes can handle difficult help-seeking interactions. In this work, we study ESDS evaluation under worst-case interactions, where seekers are hard to help due to low engagement, resistance, limited self-disclosure, emotional volatility, or rigid negative interpretations. We first conduct an expert simulation study with eight experienced counselling professionals, who simulate difficult seekers, interact with existing Chinese ESDSes, provide scale ratings, and participate in semi-structured interviews. Based on this study, we derive worst-case seeker behaviours and identify key limitations of current systems. We then propose a worst-case evaluation framework consisting of an LLM-based worst-case seeker simulator and four worst-case-oriented metrics: Deep Emotional Understanding, Guided Exploration, Balanced Emotional Support, and Authentic and Grounded Support. Evaluating 17 systems, we find that nearly all models suffer substantial performance drops under worst-case interactions. Large general-purpose LLMs are generally more robust than specialised ESDSes, but even the strongest models struggle to sustain engagement and improve seekers' emotional states. Finally, we show that worst-case simulation can also generate useful training data, improving the robustness of smaller models.

Evaluation of SiPM based Optical Receiver for Low Energy Devices

Silicon Photomultipliers (SiPMs) are photon-counting detectors with great potential to improve the sensitivity of optical receivers. However, its recovery time and dark count rate could limit its dynamic range that effectively detects single photons, hence limiting the sensitivity. Recent studies of SiPMs in communication focus on the speed rather than the power consumption of the receiver. The gain and bandwidth of the designed receiver circuit to read out SiPM output are much higher than required for the target data rate. Additionally, the SiPM experiments for optical communication are performed using an offline method which uses instruments including oscilloscopes and personal computers to process chunks of the transmitted data. In this work, we have developed an embedded realtime system FPGA-based platform to evaluate a commercially available 1 mm-sq SiPM. Moreover, simulations for investigating the relationship between the electrical bandwidth of the receiver circuit and the target data rate to achieve a target Bit Error Rate (BER) are established. The experimental results show that the implemented real-time system achieves a Bit Error Rate (BER) of 1E-3 with less than 5 pW of incident optical power with an average of 11.35 photons per bit at 100 kbps under 620 nm LED, approaching the Poisson limit. It was found that the minimum receiver electrical bandwidth required to maintain the Poisson limit is 50 times higher than the target data rate. The analysis of the minimum receiver bandwidth in photon counting and BER enables the potential future adoption of this receiver technology in high sensitivity applications in underwater and visible light communications, especially for low energy devices.