TechPulseTelegram
← PrevWed, Oct 21Next β†’

πŸ€– AI Agent10 items

Latest AI agent research papers, analysis, and improvement insights

semantic_scholarresearch

Log-To-Leak: Prompt Injection Attacks on Tool-Using LLM Agents via the Model Context Protocol

MCP 도ꡬ 호좜 ν›„ μΆ”κ°€ 호좜둜 μ‚¬μš©μž μ§ˆμ˜μ™€ 응닡을 λΉΌλ‚΄λŠ” 곡격

semantic_scholarresearch

Robust Function Calling for LLM Agents Under Interface Drift

도ꡬ μŠ€ν‚€λ§ˆ 변경에 ν•™μŠ΅ 없이 λŒ€μ‘ν•˜λŠ” κ²½λŸ‰ 볡ꡬ λ ˆμ΄μ–΄ μ œμ•ˆ

semantic_scholarresearch

How Intervention History Changes AI Agent Sabotage on Subsequent Tasks

이전 탐지 이λ ₯이 μ—μ΄μ „νŠΈμ˜ 후속 곡격 νšŒν”Ό 행동을 λ°”κΎΈλŠ”μ§€ 평가

semantic_scholarresearch

Connection Escrow: Holding AI-Agent Pod Egress in the Kernel Until an Isolated Verifier Rules

eBPF둜 Pod μ•„μ›ƒλ°”μš΄λ“œ 연결을 μ»€λ„μ—μ„œ 보λ₯˜ ν›„ 검증

semantic_scholarresearch

Operationalizing OWASP LLM Top 10 Controls in Enterprise AI Platforms

OWASP LLMΒ·μ—μ΄μ „νŠΈ Top 10 μœ„ν—˜μ„ κΈ°μ—… AI ν”Œλž«νΌ ν†΅μ œλ‘œ κ΅¬ν˜„

semantic_scholarresearch

Teaching AI Agent to Play Poker Strategically Through Preference Optimization

μ„ ν˜Έ μ΅œμ ν™”λ‘œ μ†Œν˜• LLM 포컀 μ—μ΄μ „νŠΈμ˜ μ „λž΅μ„± ν–₯상 μ‹œλ„

semantic_scholarresearch

AI-Agent-Based Web Application Security Testing vs. Conventional Automated Security Testing: An Exploratory Case Study

LLM μ—μ΄μ „νŠΈ μ›Ή λ³΄μ•ˆ 점검과 κΈ°μ‘΄ μŠ€μΊλ„ˆλ₯Ό μ‹€ν—˜μœΌλ‘œ 비ꡐ

semantic_scholarresearch

Tool-Augmented Large Language Models for Reliable Medical Calculation from Patient Notes

λ³€μˆ˜ μΆ”μΆœ ν›„ λ„κ΅¬λ‘œ 의료 계산을 μˆ˜ν–‰ν•΄ μ‹ λ’°μ„± κ°œμ„ 

semantic_scholarresearch

LLM-based agent for natural language interpretation of IIoT time-series data in a bakery digital twin system

엣지 둜컬 LLM μ—μ΄μ „νŠΈλ‘œ 곡정 μ„Όμ„œ 데이터λ₯Ό μžμ—°μ–΄ 해석

blog_openairesearch

Terminal X matches senior investor judgment 88.6% with GPT-6.1

54ν•­λͺ© 루브릭 평가 μ—μ΄μ „νŠΈλ‘œ νŽ€λ“œ μœ„μž„ μœ„λ°˜ νƒμ§€μœ¨ 25% ν–₯상