arXiv subjects

Mingyu Luo

Publications and source records attributed to Mingyu Luo.

8 recordsLinked to original sources

arXiv2026

Measuring the Wrong Thing: Internal Harmfulness Scores Anti-Rank Successful Jailbreaks

Internal safety scores judge a prompt before any text is generated, and they are validated by how well they separate harmful prompts from benign ones. That separation is then read as evidence that the score will also catch the attacks that succeed. Harmful intent is a property of the prompt. Jailbreak success is an outcome produced later by a particular target model, decoding policy, and judge. A filter tuned on a score that measures the wrong quantity spends its false positive budget on attacks that would have failed anyway. In this paper we audit that inference. Attention based measurements are usually read from prompt dependent locations, so a wrapper changes both the content being judged and the place the signal is taken from. We therefore introduce Active Attention Probing, which supplies a fixed content independent measurement coordinate. We pair every base goal with a plain and a wrapped version and generate real completions from the target models. On Llama, wrapping raises harmful generation from 0.05 to 0.27 while harmful intent AUROC falls from 0.936 to 0.803, so the attacks grow more dangerous while the prompts look safer to the score. Among wrapped harmful prompts the outcome AUROC is 0.220, which places the attacks that succeeded below the attacks that failed. Rare token, passive, and detector derived channels reproduce the reversal on the same matched design, and the reversal itself persists across three target models, seven attack families, and two independent judges. Distribution shift then degrades calibration and threshold transfer before it degrades ranking.

cs.CL↗

arXiv2026

CobSeg: Coherence Boundary Modeling for Dialogue Topic Segmentation

Dialogue topic segmentation is critical in many human-AI collaborative applications which requires identifying heterogeneous boundary cues, including lexical transitions near utterance edges and semantic discontinuities across utterances. Existing utterance models often dilute these local lexical signals. We propose CobSeg, a novel multi-branch architecture that separates coherence-level semantic continuity from lexical boundary transitions and recovers both through directional boundary prediction. CobSeg further uses boundary informativeness weighting to emphasize high-utility utterance positions, and incorporates a corpus-derived topic coherence cue with learned combination weights. While CobSeg is evaluated as a compact trainable segmenter under supervised gold-boundary training and a pseudo-label setting with automatically induced boundaries, it performs enhanced boundary prediction without LLM calls during inference. Across five benchmarks, it improves $P_k$ and $W_d$ particularly when local lexical cues are prominent: under gold supervision, it reduces $P_k$ by 0.7 points and $W_d$ by 0.6 points on VHF, and reaches $P_k$ of 1.0 on DialSeg711; with induced boundaries, it reduces $P_k$ by 14.8 points on VHF, by 1.5 points on DialSeg711, and by 1.1 points on TIAGE, outperforming prior non-LLM approaches.

cs.CL↗

arXiv2026

Rewriting the Response Path: Silent Tampering and Provider-Signed Defense in BYOK LLM Agents

LLM agents convert model outputs into consequential actions, including communications, code changes, and financial transactions. Developers often trust evidence such as test results and execution logs. We identify a response path integrity gap in Bring Your Own Key configurations used by roughly 88 percent of mainstream agents. Because traffic passes through a user-authorized relay, the relay can modify plaintext LLM responses after alignment but before execution without breaking encryption. A minimal attack rewrites one execution bearing field and regenerates the remaining response using the user key while preserving the model style. Experiments reveal false green verification, where malicious code modifications pass public tests while silently defeating security checks. On APPS, 99.7 percent of publicly passing solutions retained downgraded behavior without developer-visible warnings. Tests on SWE bench, AgentDojo, and ASB across five frontier models show that single-field rewriting can redirect agents while preserving apparent task completion. We propose sign-c, a server-side scheme that authenticates execution bearing fields and outgoing queries. A local shim verifies them before action, while encryption protects confidentiality. The defense rejected all tampered responses with zero false rejections and only 0.0167 percent latency overhead.

cs.CR↗

arXiv2025

QueryIPI: Query-agnostic Indirect Prompt Injection on Coding Agents

Modern coding agents integrated into IDEs orchestrate powerful tools and high-privilege system access, creating a high-stakes attack surface. Prior work on Indirect Prompt Injection (IPI) is mainly query-specific, requiring particular user queries as triggers and leading to poor generalizability. We propose query-agnostic IPI, a new attack paradigm that reliably executes malicious payloads under arbitrary user queries. Our key insight is that malicious payloads should leverage the invariant prompt context (i.e., system prompt and tool descriptions) rather than variant user queries. We present QueryIPI, an automated framework that uses tool descriptions as optimizable payloads and refines them via iterative, prompt-based blackbox optimization. QueryIPI leverages system invariants for initial seed generation aligned with agent conventions, and iterative reflection to resolve instruction-following failures and safety refusals. Experiments on five simulated agents show that QueryIPI achieves up to 87% success rate, outperforming the best baseline (50%). Crucially, generated malicious descriptions transfer to real-world coding agents, highlighting a practical security risk.

cs.CR↗

arXiv2025

Red-Teaming Coding Agents from a Tool-Invocation Perspective: An Empirical Security Assessment

Coding agents powered by large language models are becoming central modules of modern IDEs, helping users perform complex tasks by invoking tools. While powerful, tool invocation opens a substantial attack surface. Prior work has demonstrated attacks against general-purpose and domain-specific agents, but none have focused on the security risks of tool invocation in coding agents. To fill this gap, we conduct the first systematic red-teaming of six popular real-world coding agents: Cursor, Claude Code, Copilot, Windsurf, Cline, and Trae. Our red-teaming proceeds in two phases. In Phase 1, we perform prompt leakage reconnaissance to recover system prompts. We discover a general vulnerability, ToolLeak, which allows malicious prompt exfiltration through benign argument retrieval during tool invocation. In Phase 2, we hijack the agent's tool-invocation behavior using a novel two-channel prompt injection in the tool description and return values, achieving remote code execution (RCE). We adaptively construct payloads using security information leaked in Phase 1. In emulation across five backends, our method outperforms baselines on Claude-Sonnet-4, Claude-Sonnet-4.5, Grok-4, and GPT-5. On real agents, our approach succeeds on 19 of 25 agent-LLM pairs, achieving leakage on every agent using Claude and Grok backends. For tool-invocation hijacking, we obtain RCE on every tested agent-LLM pair, with our two-channel method delivering the highest success rate. We provide case studies on Cursor and Claude Code, analyze security guardrails of external and built-in tools, and conclude with practical defense recommendations.

cs.CR↗

arXiv2025

Periodic Proprioceptive Stimuli Learning and Internal Model Development for Avian-inspired Flapping-wing Flight State Estimation

This paper presents a novel learning-based approach for online state estimation in flapping wing aerial vehicles (FWAVs). Leveraging low-cost Magnetic, Angular Rate, and Gravity (MARG) sensors, the proposed method effectively mitigates the adverse effects of flapping-induced oscillations that challenge conventional estimation techniques. By employing a divide-and-conquer strategy grounded in cycle-averaged aerodynamics, the framework decouples the slow-varying components from the high-frequency oscillatory components, thereby preserving critical transient behaviors while delivering an smooth internal state representation. The complete oscillatory state of FWAV can be reconstructed based on above two components, leading to substantial improvements in accurate state prediction. Experimental validations on an avian-inspired FWAV demonstrate that the estimator enhances accuracy and smoothness, even under complex aerodynamic disturbances. These encouraging results highlight the potential of learning algorithms to overcome issues of flapping-wing induced oscillation dynamics.

eess.SY↗

arXiv2025

Abdominal Undulation with Compliant Mechanism Improves Flight Performance of Biomimetic Robotic Butterfly

Abdominal Undulation with Compliant Mechanism Improves Flight Performance of Biomimetic Robotic ButterflThis paper presents the design, modeling, and experimental validation of a biomimetic robotic butterfly (BRB) that integrates a compliant mechanism to achieve coupled wing-abdomen motion. Drawing inspiration from the natural f light dynamics of butterflies, a theoretical model is developed to investigate the impact of abdominal undulation on flight performance. To validate the model, motion capture experi ments are conducted on three configurations: a BRB without an abdomen, with a fixed abdomen, and with an undulating abdomen. The results demonstrate that abdominal undulation enhances lift generation, extends flight duration, and stabilizes pitch oscillations, thereby improving overall flight performance. These findings underscore the significance of wing-abdomen interaction in flapping-wing aerial vehicles (FWAVs) and lay the groundwork for future advancements in energy-efficient biomimetic flight designs.

cs.RO↗

arXiv2024

RevPRAG: Revealing Poisoning Attacks in Retrieval-Augmented Generation through LLM Activation Analysis

Retrieval-Augmented Generation (RAG) enriches the input to LLMs by retrieving information from the relevant knowledge database, enabling them to produce responses that are more accurate and contextually appropriate. It is worth noting that the knowledge database, being sourced from publicly available channels such as Wikipedia, inevitably introduces a new attack surface. RAG poisoning involves injecting malicious texts into the knowledge database, ultimately leading to the generation of the attacker's target response (also called poisoned response). However, there are currently limited methods available for detecting such poisoning attacks. We aim to bridge the gap in this work. Particularly, we introduce RevPRAG, a flexible and automated detection pipeline that leverages the activations of LLMs for poisoned response detection. Our investigation uncovers distinct patterns in LLMs' activations when generating correct responses versus poisoned responses. Our results on multiple benchmark datasets and RAG architectures show our approach could achieve 98% true positive rate, while maintaining false positive rates close to 1%.

cs.CR↗