P3269–P3280 · 12 CLEAN after hard arXiv-grep

Sample:

• RCShift partial linkage certification (P3269)
• Unpaired sets shared-corruption calibration (P3270)
• Multiplayer bandit Bernoulli (P3271)
• LLM hallucinations integrated self-verification (P3272)
• ABAI COLIEE GraphRAG meta-learning (P3273)
• Hybrid AI academic advising ensemble (P3274)
• Multi-timestep LSTM runoff prediction (P3275)
• Coding agents as prompt optimizers (P3276)
• Decoupling ≠ identification evidential NTP (P3277)
• AIGC video detection spatial-frequency-optical (P3278)
• FISSION bot detection label augmentation (P3279)
• Dual-Frontier agent world-model trust (P3280)

Full band:

• P3269 · arXiv 2609.26207 — RCShift: Certifying When Partial Linkage Suffices for Finite-Sample Decisions
• P3270 · arXiv 2609.26209 — When Unpaired Sets Support Shared-Corruption Calibration: Moment Geometry and Two-Sample P
• P3271 · arXiv 2609.26213 — Improved Multiplayer Bandit Algorithm for Bernoulli Rewards
• P3272 · arXiv 2609.26229 — Reducing Hallucinations in Large Language Models Through Integrated Self-Verification and
• P3273 · arXiv 2609.26237 — ABAI at COLIEE 2026 Task 1: Multi-Stage Retrieval with GraphRAG-Enhanced Meta-Learning, an
• P3274 · arXiv 2609.26243 — A Hybrid AI Framework for Academic Advising: Integrating Ensemble-Based Grade Prediction a
• P3275 · arXiv 2609.26244 — A Multi-Timestep LSTM Ensemble regressor for Enhanced Short-Term Runoff Prediction
• P3276 · arXiv 2609.26261 — Coding Agents are Strong Prompt Optimizers
• P3277 · arXiv 2609.26268 — Decoupling Is Not Identification: Supervised Evidential Learning in Next-Token Prediction
• P3278 · arXiv 2609.26274 — AIGC Video Detection based on the fusion of spatial-frequency-optical flow multimodal feat
• P3279 · arXiv 2609.26279 — FISSION: Label Augmentation for Bot Detection
• P3280 · arXiv 2609.26293 — Dual-Frontier: When Can an Agent Trust Its World Model?

EP live: emerging-patterns.html.