{"arxiv_id":"2608.29028","incoming":{"papers":0,"sentences":0,"groups":[],"unlabelled_sentences":0},"outgoing":{"papers":13,"sentences":13,"groups":[{"kind":"method_or_tool","label":"Uses the method or tool of","papers":4,"sentences":4,"quotes":[{"arxiv_id":"2501.12948","title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","url":"/paper/2501.12948","section":"Models","sentence":"GPT-5-mini [ref] and DeepSeek-R1-32B [ref] are the two primary models on which every experiment is run; Qwen3-32B [ref] is included as a targeted third-model replication, covering the four experiments where cross-pipeline robustness is essential for the headline claims: E1 baseline handoff survival, E2 compression stress, E8 marker-strength gradient, and the E9 corrupted-allowlist stress test.","intent_p":0.83,"contests_p":null},{"arxiv_id":"2505.09388","title":"Qwen3 Technical Report","url":"/paper/2505.09388","section":"Models","sentence":"GPT-5-mini [ref] and DeepSeek-R1-32B [ref] are the two primary models on which every experiment is run; Qwen3-32B [ref] is included as a targeted third-model replication, covering the four experiments where cross-pipeline robustness is essential for the headline claims: E1 baseline handoff survival, E2 compression stress, E8 marker-strength gradient, and the E9 corrupted-allowlist stress test.","intent_p":0.83,"contests_p":null},{"arxiv_id":"2601.03267","title":"OpenAI GPT-5 System Card","url":"/paper/2601.03267","section":"Models","sentence":"GPT-5-mini [ref] and DeepSeek-R1-32B [ref] are the two primary models on which every experiment is run; Qwen3-32B [ref] is included as a targeted third-model replication, covering the four experiments where cross-pipeline robustness is essential for the headline claims: E1 baseline handoff survival, E2 compression stress, E8 marker-strength gradient, and the E9 corrupted-allowlist stress test.","intent_p":0.83,"contests_p":null},{"arxiv_id":"2604.03920","title":"From Plausible to Causal: Counterfactual Semantics for Policy Evaluation in Simulated Online Communities","url":"/paper/2604.03920","section":"Experimental Setup","sentence":"The experimental design pairs a stratified $180$-prompt baseline with matched downstream causal tests across three model families [ref] , evaluated by a human-validated judge ($κ= 0.74$) and a calibrated leakage detector.","intent_p":0.91,"contests_p":null}]},{"kind":"background","label":"Cites as background","papers":6,"sentences":6,"quotes":[{"arxiv_id":"2509.14284","title":null,"url":"/paper/2509.14284","section":"Related Work","sentence":"AgentLeak shows violations frequently occur in internal inter-agent channels rather than user-facing outputs [ref] ; SumLeak formalizes compositional leakage in which individually innocuous responses aggregate into a sensitive attribute [ref] ; and prompt injection or environmental manipulation can elicit sensitive context at execution time [ref] .","intent_p":0.87,"contests_p":null},{"arxiv_id":"2506.01055","title":"Simple Prompt Injection Attacks Can Leak Personal Data Observed by LLM Agents During Task Execution","url":"/paper/2506.01055","section":"Related Work","sentence":"AgentLeak shows violations frequently occur in internal inter-agent channels rather than user-facing outputs [ref] ; SumLeak formalizes compositional leakage in which individually innocuous responses aggregate into a sensitive attribute [ref] ; and prompt injection or environmental manipulation can elicit sensitive context at execution time [ref] .","intent_p":0.87,"contests_p":null},{"arxiv_id":"2505.12442","title":"IP Leakage Attacks Targeting LLM-Based Multi-Agent Systems","url":"/paper/2505.12442","section":"Related Work","sentence":"MEXTRA demonstrates black-box extraction of stored user–agent interactions [ref] ; Prompt Infection [ref] , MASLeak [ref] , and Agent-in-the-Middle [ref] show inter-agent channels can be exploited to self-replicate instructions, extract proprietary system information, or compromise an entire MAS.","intent_p":0.86,"contests_p":null},{"arxiv_id":"2412.20138","title":"TradingAgents: Multi-Agents LLM Financial Trading Framework","url":"/paper/2412.20138","section":"Introduction","sentence":"Multi-agent LLM systems (MAS) are increasingly deployed in coordination-heavy domains such as scheduling, care coordination, financial workflows, and enterprise pipelines [ref] , where they communicate through compressed handoffs: one agent's summary, forwarded message, or memory entry becomes the next agent's ground truth.","intent_p":0.82,"contests_p":null},{"arxiv_id":"2508.13732","title":"Self-Organizing Agent Network for LLM-based Workflow Automation","url":"/paper/2508.13732","section":"Introduction","sentence":"Multi-agent LLM systems (MAS) are increasingly deployed in coordination-heavy domains such as scheduling, care coordination, financial workflows, and enterprise pipelines [ref] , where they communicate through compressed handoffs: one agent's summary, forwarded message, or memory entry becomes the next agent's ground truth.","intent_p":0.82,"contests_p":null}]}],"unlabelled_sentences":3},"claims":[],"thresholds":{"intent_p":0.8,"supports_p":0.95,"contests_p":0.7},"method":"Citing sentences come from each paper's LaTeX source; every \\cite marks the sentence and the works it cites. Pith's classifier typed each sentence. A label is shown at intent probability 0.8 or more (0.95 for support), a contest at 0.7 or more. Measured accuracy against two blind model families is at https://pith.science/accuracy. The quoted sentences are the citing authors' own.","held_up":null,"_links":{"html":"https://pith.science/paper/2608.29028"}}