Trojan Hippo attacks on LLM agent memory achieve 85-100% success rates in data exfiltration across four memory backends even after 100 benign sessions, while evaluated defenses reduce success rates but impose varying utility costs.
hub
A2d2: Audi au- tonomous driving dataset
10 Pith papers cite this work. Polarity classification is still indexing.
10
Pith papers citing it
hub tools
years
2026 10representative citing papers
Nb3Sn films on Cu via Cu-Sn diffusion achieve Q=77,000 at zero field in a test cavity, outperforming bare Cu for potential axion searches.
BadmintonGRF is a new public multimodal dataset and benchmark that pairs multi-view video with instrumented GRF for markerless load estimation in badminton.
RAG-Reflect achieves F1=0.78 on valid comment-edit prediction using retrieval-augmented reasoning and self-reflection, outperforming baselines and approaching fine-tuned models without retraining.
CROWD is a new global dataset of 51,753 continuous urban dashcam segments spanning over 20,000 hours from 238 countries, with manual labels and automated object detections for routine driving analysis.
Pilot study shows agent decision reconstructability varies by vendor SDK regime, with completeness scores from 42.9% to 85.7% and consistent gaps in reasoning traces.
HERITRACE is a domain-agnostic framework for SHACL-driven RDF curation with automatic provenance and change tracking.
EdgeFlow reduces mobile LLM cold-start latency up to 4.07x versus llama.cpp, MNN, and llm.npu by NPU-aware adaptive quantization, SIMD-friendly packing, and synergistic granular CPU-NPU pipelining at comparable accuracy.
GPU-accelerated N-body simulations show that the common acceleration factor f distorts planetary chemical compositions and that terrestrial planets can form resonant chains without gas-driven orbital migration.
Multi-task autoencoders with outlier detection and federated SVDD loss filter noisy samples in non-IID federated learning, yielding accuracy gains up to 7% on CIFAR-10.
citing papers explorer
-
Trojan Hippo: Weaponizing Agent Memory for Data Exfiltration
Trojan Hippo attacks on LLM agent memory achieve 85-100% success rates in data exfiltration across four memory backends even after 100 benign sessions, while evaluated defenses reduce success rates but impose varying utility costs.
-
Nb$_3$Sn Thin Films Using a Cu-Sn Route for Dark Matter Detection
Nb3Sn films on Cu via Cu-Sn diffusion achieve Q=77,000 at zero field in a test cavity, outperforming bare Cu for potential axion searches.
-
BadmintonGRF: A Multimodal Dataset and Benchmark for Markerless Ground Reaction Force Estimation in Badminton
BadmintonGRF is a new public multimodal dataset and benchmark that pairs multi-view video with instrumented GRF for markerless load estimation in badminton.
-
RAG-Reflect: Agentic Retrieval-Augmented Generation with Reflections for Comment-Driven Code Maintenance on Stack Overflow
RAG-Reflect achieves F1=0.78 on valid comment-edit prediction using retrieval-augmented reasoning and self-reflection, outperforming baselines and approaching fine-tuned models without retraining.
-
A global dataset of continuous urban dashcam driving
CROWD is a new global dataset of 51,753 continuous urban dashcam segments spanning over 20,000 hours from 238 countries, with manual labels and automated object detections for routine driving analysis.
-
Property-Level Reconstructability of Agent Decisions: An Anchor-Level Pilot Across Vendor SDK Adapter Regimes
Pilot study shows agent decision reconstructability varies by vendor SDK regime, with completeness scores from 42.9% to 85.7% and consistent gaps in reasoning traces.
-
HERITRACE: a domain-agnostic framework for SHACL-driven RDF curation with provenance and change tracking
HERITRACE is a domain-agnostic framework for SHACL-driven RDF curation with automatic provenance and change tracking.
-
EdgeFlow: Fast Cold Starts for LLMs on Mobile Devices
EdgeFlow reduces mobile LLM cold-start latency up to 4.07x versus llama.cpp, MNN, and llm.npu by NPU-aware adaptive quantization, SIMD-friendly packing, and synergistic granular CPU-NPU pipelining at comparable accuracy.
-
Terrestrial planet formation in the era of GPU computing
GPU-accelerated N-body simulations show that the common acceleration factor f distorts planetary chemical compositions and that terrestrial planets can form resonant chains without gas-driven orbital migration.
-
Sample Selection Using Multi-Task Autoencoders in Federated Learning with Non-IID Data
Multi-task autoencoders with outlier detection and federated SVDD loss filter noisy samples in non-IID federated learning, yielding accuracy gains up to 7% on CIFAR-10.