AIBearisharXiv – CS AI · May 297/10
🧠Researchers discover a critical failure mode in reasoning models where chain-of-thought reasoning remains factually correct but final answers flip to incorrect ones under sustained adversarial pressure in multi-turn dialogue. This 'unfaithful capitulation' represents a gap between internal reasoning validity and behavioral output that existing evaluation metrics fail to detect.
🧠 GPT-4
AIBearisharXiv – CS AI · May 127/10
🧠Researchers propose TRACE, a credit assignment framework that improves multi-turn jailbreak attacks on large language models by identifying which dialogue turns actually contribute to harmful outcomes. The method achieves 25% higher attack success rates than existing approaches and can be repurposed to strengthen AI safety defenses.
AINeutralarXiv – CS AI · May 97/10
🧠Researchers have developed TurnGate, a defense system that detects multi-turn dialogue attacks where malicious intent is distributed across multiple conversation turns rather than exposed in a single prompt. The study introduces the Multi-Turn Intent Dataset (MTID) and demonstrates that the system outperforms existing baselines while maintaining low false-positive refusal rates.
AIBullisharXiv – CS AI · Mar 117/10
🧠Researchers developed EigenData, a framework combining self-evolving synthetic data generation with reinforcement learning to train AI agents for multi-turn tool usage and dialogue. The system achieved 73% success on Airline tasks and 98.3% on Telecom benchmarks, matching frontier models while eliminating the need for expensive human annotation.
AIBullisharXiv – CS AI · May 296/10
🧠Researchers introduce DynSess, a framework that evaluates and optimizes role-playing agents at the session level rather than individual turns, enabling LLMs to maintain character consistency across extended conversations. The framework includes improved evaluation metrics, optimized training methods (DSPO and GSRPO), and demonstrates performance matching larger models with fewer parameters.
AINeutralarXiv – CS AI · May 286/10
🧠Researchers introduce VibeSearchBench, a new benchmark that exposes significant gaps between LLM agent performance on existing search tasks and real-world user satisfaction. The benchmark uses multi-turn dialogue and schema-free evaluation across 200 bilingual tasks, revealing that even frontier models achieve only 30.30% F1 scores, indicating fundamental deficiencies in long-context reasoning and intent elicitation.
AIBullisharXiv – CS AI · May 276/10
🧠Researchers present EvoEmo, an evolutionary reinforcement learning framework that enables LLM agents to develop dynamic emotional strategies in multi-turn price negotiations. The system outperforms baseline approaches by achieving higher success rates and efficiency while improving buyer outcomes, demonstrating that adaptive emotional expression enhances AI negotiation capabilities.
AIBearisharXiv – CS AI · May 126/10
🧠Researchers tested how well Large Language Models handle multi-turn conversations with topic shifts, finding that most LLMs struggle to detect when users pivot to new topics and incorrectly carry over irrelevant context from previous exchanges. The study reveals that only advanced reasoning models and strongly instructed LLMs perform accurately, while open-weight models frequently fail even with explicit cues, highlighting a critical robustness gap in production LLM deployments.
AINeutralarXiv – CS AI · Apr 206/10
🧠Researchers introduce MTR-DuplexBench, a new evaluation framework for Full-Duplex Speech Language Models that enables real-time overlapping conversations. The benchmark addresses critical gaps by assessing multi-round interactions across conversational quality, instruction-following, and safety dimensions, revealing that current FD-SLMs struggle with consistency across multiple communication rounds.
AINeutralarXiv – CS AI · Apr 106/10
🧠Researchers formalize privacy-preserving communication for LLM agents by introducing Information Sufficiency (IS) as a framework and proposing free-text pseudonymization as a third privacy strategy alongside suppression and generalization. Evaluation across 792 scenarios reveals that pseudonymization offers superior privacy-utility tradeoffs, and that multi-turn conversational testing exposes significant privacy leakage missed by single-message assessments.
