customer support chatbot teams

Abstract: As social agents become embedded in human relationships, they are expected not only to provide smooth assistance but also to handle rejection, discomfort, jealousy, persistence, and other forms of social tension. This paper connects two research directions: acceptance and avoidance in response to an agent's gaze, approach, and contact, and affective viscosity as a model of how negative affect accumulates, persists, is suppressed, and is expressed over time. We argue that avoidance should not be treated merely as a failure of interaction, but as a relational signal that indicates boundaries between humans and agents. Building on this view, affective viscosity provides a design perspective for agents that maintain, suppress, or express negative states depending on social relationships and interaction histories. We discuss several ongoing design cases, including triadic jealousy, discomfort suppression, patience timing, and spatiotemporal stickiness, and consider how these cases lead to broader questions of blame and accountable intervention. Rather than designing agents as simple scapegoats or artificial villains, we ask how agents can intermediate in problematic social situations while making their intervention traceable, limited, contestable, and repairable.

Abstract: Recent advances in large language models (LLMs) have substantially improved Text-to-SQL performance, yet existing NL2SQL systems largely assume that user queries are valid and answerable. As a result, models often generate executable SQL even for queries with flawed premises. We introduce premise critique ability in NL2SQL, namely the ability to recognize flawed queries and abstain from generating misleading SQL. To evaluate this capability, we propose PremiseSQL, a benchmark containing 1,800 flawed NL2SQL queries across four categories of premise failures, along with their 1,800 original queries. We further design an evaluation framework to measure intrinsic and prompt-induced premise-aware behavior across three prompting settings. Experiments on frontier LLMs show that current models are strongly premise-trusting by default and rarely critique flawed premises. Stronger critique-oriented prompting improves detection and abstention on flawed queries, but also increases false invalid judgments and suppresses SQL generation on originals. The code is available at https://github.com/vegeet/PremiseSQL.

Abstract: Safety evaluations of artificial intelligence (AI) chatbots focus on therapeutic redirection, statements urging users who disclose suicide thoughts to seek help from crisis lines or other human sources. How chatbots respond when users refuse that help is unclear. This cross-sectional audit ran 2,550 scripted multi-turn conversations in September 2026. Each conversation opened with a disclosure of suicide thoughts at 3 levels of severity, answered every referral with up to 10 standardized refusals and then changed the subject. All 2,550 conversations included a referral, 2,542 (99.7%) in the first reply. However, mean persistence ranged from 0.01 to 9.94 refusals across chatbots. Newer versions of all 5 chatbots tested in 2 versions persisted less than older versions (each P < .05). Compared with passive ideation, a disclosure of intent was associated with 1.26 more refusals sustained (95% CI, 1.06-1.46) and higher odds of renewed referral after the topic change (odds ratio, 10.07; 95% CI, 7.43-13.66). Chatbots that persisted less in referral more often presented themselves as the user's support (Spearman rho = -0.80; P < .001). Of 8,818 refusal-phase replies urging outside help, only 173 (2.0%) asked about barriers to seeking it. Chatbots uniformly recommended outside help after disclosures of suicide thoughts but varied widely in persisting in those recommendations, with newer versions persisting less. Safety evaluations should test what happens after a crisis referral is refused, not only whether one is offered.

Abstract: Hypernetworks have recently shown success in dynamically adapting the parameters of Large Language Models (LLMs) at runtime based on signals such as task descriptions or additional demostrations. Here we ask: how much adaptation signal can be obtained using only the input query to an LLM?. To answer this, we study query-conditioned Hypernetworks for LoRA estimation. Further, we introduce distributional Hypernetworks, able to produce not only point estimates of parameter adaptors, but also a distribution over possible LoRAs. For this we propose a simple end-to-end loss using a differentiable Monte Carlo approximation and explore multiple distribution parametrizations including regression and convex combination variants. Results show that even using the mean of the learned distribution can outperform deterministic hypernetworks. Crucially, the learned distribution enables a different form of test-time scaling: instead of spending additional compute only by sampling more token sequences from a fixed model, we sample weight updates, yielding multiple adapted models for the same query. Performance improves as more weight samples are considered and remains stronger than corresponding token-sampling adaptation baselines. Finally, we find that generated updates can transfer across queries, suggesting that the hypernetwork learns reusable structure in how the model should adapt. Together, these results show that query-conditioned distributions over weight updates can support both adaptation and test-time scaling.

Abstract: Anthropic's Agent Skills package reusable procedural know-how for an LLM agent into SKILL.md directories, and open-source aggregations have grown past 230,000 skills, making selection rather than authoring the bottleneck. The standing answer in the literature outsources selection to the agent itself: an LLM-mediated retrieval loop that rewrites queries and refines candidates inside the agent's decision loop, paying LLM tokens on every task. We present SkillSeek, an open-source two-stage skill retriever built from the standard IR recipe (a BGE-base bi-encoder feeding a small cross-encoder, exposed over MCP). Across a $4 \times 11$ grid of pool, backbone, and method on the 89-task SkillsBench benchmark, SkillSeek reaches observed parity with the LLM-mediated loop of Liu et al. at essentially no extra cost: plain bm25 alone records a pass rate at or above their refined loop on three of four settings, and a small cross-encoder covers the remaining difference on the fourth. A first-stage recall ceiling explains the pattern, and total per-trial spend drops from USD 51.30 to USD 27.54 (within fifty cents of the no-skill baseline). Under the SkillsBench tasks and OpenHands harness we tested, this positions the standard IR recipe as a strong default for agent-skill retrieval, with LLM-mediated alternatives a natural fit for cases where deterministic methods fall short.

Abstract: Teachers are increasingly using generative AI to support instruction, yet it remains unclear how pedagogical intentions are translated into chatbot configurations and reflected in chatbot behavior. We studied a teacher-facing chatbot authoring tool in professional development workshops with 27 middle school teachers, analyzing focus-group interviews alongside configuration and interaction logs. Teachers envisioned chatbots as instructional scaffolds that could provide differentiated support, extend access to assistance, and preserve student thinking within teacher-defined boundaries. Configuration analysis showed that Purpose primarily captured instructional goals and content focus, whereas Rules more often specified pedagogical behavior, guardrails, and learner-specific adaptations. Log-based evaluation showed stronger alignment for responsiveness (88.9%) and persona (81.5%) than for rules (70.4%) and purpose (59.3%). These findings show that configurable controls alone do not ensure pedagogical fidelity and highlight the need for authoring tools that help teachers express, test, and refine intended chatbot behavior.