Post-implant telephone follow-up requires an agent to respond naturally, complete a structured standard operating procedure (SOP), and avoid unsupported medical advice within the same call. We present SAFER, a live speech-first system for Chinese neuromodulation follow-up that exposes three capabilities in one demonstration: real-time voice interaction with barge-in, an inspectable in-call SOP trace, and guarded handling of unscripted medical or device questions. In the dual-lane architecture, user-facing speech does not wait for asynchronous state commitment, while questions assigned to the guarded branch receive fixed non-generative deflection phrases; callback-eligible clinical or safety subclasses also create a human-follow-up record. Controlled studies report a median end-to-end voice latency of 1.07 seconds, a 0.983 transcript-level completeness-accuracy field score, and 23/30 joint routing, correctness, coverage, and grounding successes on independently expert-reviewed synthetic gold-ANSWER queries. The hosted demo lets reviewers conduct a call and inspect the evolving in-call record, available evidence text, latency, and patient-facing routing behavior. These results characterize a simulated prototype; they do not establish clinical safety or real-world robustness.

With the widespread adoption of open-source code language models (code LMs), intellectual property (IP) protection has become an increasingly critical concern. While current watermarking techniques have the potential to identify the code LM to protect its IP, they have limitations when facing the more practical and complex demand, i.e., offering the individual user-level tracing in the black-box setting. This work presents CLMTracing, a black-box code LM watermarking framework employing the rule-based watermarks and utility-preserving injection method for user-level model tracing. CLMTracing further incorporates a parameter selection algorithm sensitive to the robust watermark and adversarial training to enhance the robustness against watermark removal attacks. Comprehensive evaluations demonstrate CLMTracing is effective across multiple state-of-the-art (SOTA) code LMs, showing significant harmless improvements compared to existing SOTA baselines and strong robustness against various removal attacks.

After large models (LMs) have gained widespread acceptance in code-related tasks, their superior generative capacity has greatly promoted the application of the code LM. Nevertheless, the security of the generated code has raised attention to its potential damage. Existing secure code generation methods have limited generalizability to unseen test cases and poor robustness against the attacked model, leading to safety failures in code generation. In this paper, we propose a generalizable and robust secure code generation method SecCoder by using in-context learning (ICL) and the safe demonstration. The dense retriever is also used to select the most helpful demonstration to maximize the improvement of the generated code’s security. Experimental results show the superior generalizability of the proposed model SecCoder compared to the current secure code generation method, achieving a significant security improvement of an average of 7.20% on unseen test cases. The results also show the better robustness of SecCoder compared to the current attacked code LM, achieving a significant security improvement of an average of 7.74%. Our analysis indicates that SecCoder enhances the security of LMs in generating code, and it is more generalizable and robust.

Chinese Spelling Correction (CSC) is a task to detect and correct misspelled characters in Chinese texts. CSC is challenging since many Chinese characters are visually or phonologically similar but with quite different semantic meanings. Many recent works use BERT-based language models to directly correct each character of the input sentence. However, these methods can be sub-optimal since they correct every character of the sentence only by the context which is easily negatively affected by the misspelled characters. Some other works propose to use an error detector to guide the correction by masking the detected errors. Nevertheless, these methods dampen the visual or phonological features from the misspelled characters which could be critical for correction. In this work, we propose a novel general detector-corrector multi-task framework where the corrector uses BERT to capture the visual and phonological features from each character in the raw sentence and uses a late fusion strategy to fuse the hidden states of the corrector with that of the detector to minimize the negative impact from the misspelled characters. Comprehensive experiments on benchmarks demonstrate that our proposed method can significantly outperform the state-of-the-art methods in the CSC task.