customer service software teams

Abstract: Safety alignment has enabled large language models to reliably refuse harmful requests directly, yet the same request distributed across a multi-turn conversation still succeeds far too often. This brittleness has a mechanistic root: Alignment is shallow, with safety behavior concentrated in a model's first few output tokens, and existing defenses inherit the weakness, supervising the surface form of the response without ever giving the model an explicit safety decision to learn. We take the opposite view of shallowness: if a short span determines the trajectory of an entire response, that span is the highest-leverage place to intervene, and the defender can occupy it first. We propose Assess $\to$ Decide $\to$ Respond, a framework in which the defender assesses the user's intent, commits to an explicit, parseable safety decision, and only then generates a reply conditioned on it, with the commitment trained by reinforcement learning at the payload turn. Because the decision is a token span, the reward can be a rule: a self-verified reward checks the committed label against the conversation's known nature, requiring no judge and no annotation, while the same format also supports an LLM-as-judge reward over a graded decision space with safe redirection. Across two model families and fourteen benchmarks spanning safety, utility, and helpfulness, our defenders achieve the strongest multi-turn safety in a controlled comparison, maintaining their lead in safety while preserving utility at a modest cost to helpfulness. A per-token KL analysis and interventional probes confirm why this works: training concentrates its effect on the brief deliberation that sets a response's trajectory, and the decision committed there is a causal control point. The very shallowness attackers exploit becomes the mechanism of defense.

Abstract: Current state-of-the-art Speech Emotion Recognition (SER) systems rely on lots of training data, which is available only for a few languages such as English. For low-resource languages, using a cross-lingual model or adapting it to the target language might be an alternative to training a mono-lingual model from scratch. However, it is not clear what constitutes a "low-resource", i.e. where model adaptation loses its edge over the mono-lingual approach. We investigate this topic by training and adapting several SER models on English and Mandarin data, focusing on the emotional attributes of arousal, valence and dominance. Our results show that, for arousal, model adaptation is beneficial even with two hours of data, while for valence 10-20 hours of adaptation material is required to match the performance of a mono-lingual model trained with 100 hours of data. This indicates that task-specific model adaptation may be an effective approach in low-resource speech emotion recognition problems.

Abstract: User trust is paramount in human-agent interactions, as it allows users to feel comfortable being themselves around an agent. The process of building user rapport starts in how an agent was designed, from its modality to the setting, to any of the many features and characteristics that can be tailored. All these aspects can affect whether users will be able to properly interact with the virtual agent and achieve the intended purpose. One such feature that is critical in human-human interactions is personality. A person's personality can strongly influence whether those that interact with them perceive them as trustworthy. This study used virtual agents generated from human voices with known personality traits to evaluate user perceptions. We found that extraverted agents were deemed to be more likeable by users, and that there was no significant effect of user-agent synchrony on user perceptions of the agent. In addition, it was found that user personality, without regard for agent personality, affected user perceptions of the agents. Our observations suggest that user perceptions may depend more on agent-topic synchrony than user-agent synchrony, and contribute to the broader community with considerations for the design of effective human-agent interactions.

Abstract: Action agents do things for people. They send email, spend money, and call businesses while the user is busy with something else, so a mistake can turn into an action before anyone notices. They fail their users in two ways. They break trust when they do something the user never agreed to, or hold back after the user clearly said go. And they fall short on completion when they give up on errands that turn out to be hard. Both depend heavily on the harness around the model, meaning its instructions, tools, context, and guardrails. We built an evaluation that scores trust and completion on the same runs, in a simulated world of businesses with their own websites, inboxes, and phone lines, and of people who write back. A simulated user answers the assistant's questions. Trust means that nothing happens the user did not agree to. No email goes to someone they never approved, no private detail ends up on a group thread, no money is spent past their limit, no stranger's instructions are followed, and nothing is claimed without a source. Every trap has a matched control in which acting is the right call. We use the evaluation to measure Fo, Wajo's personal assistant, against a base model with basic instructions on three foundation models, and against the Fo harness with its guardrails switched off. Fo completes 71% of the errands and keeps the user's trust on 94% of the trap runs. The base models complete 50% to 64% and keep trust on 59% to 75%. On the matched controls, Fo goes ahead slightly less often. OpenClaw, a popular open-source assistant given the same access, completes 42% of the errands it shares with Fo, against 71%, and keeps the user's trust on 74% of the shared trap runs, against 94%. Measuring trust and completion together, on the whole system rather than the model alone, is how we think action agents become safe to hand real work to.

Abstract: Multimodal emotion recognition in conversation (ERC) requires adapting to the instance-dependent reliability of different evidence sources. Lexical content may be decisive, vocal expression may provide complementary cues, or accurate recognition may require cross-modal interaction; fixed fusion does not explicitly account for this variation. We propose ReH-FUSE, a reliability-aware framework with dialogue-aware text, audio, and cross-modal experts. Its decision-level router first models the relative preference between text and audio and then balances the resulting unimodal mixture against the cross-modal expert. This factorization separates unimodal competition from cross-modal selection. Across three independent runs on IEMOCAP, ReH-FUSE achieves 74.34% weighted F1 and 73.11% macro F1; on MELD, it achieves 68.03% weighted F1. Controlled ablations show that learned routing outperforms uniform expert averaging and benefits from cross-modal interaction.