Performance and Energy Trade-offs in Multi-Request LLM Workflows
Explore how multi-request workflows impact large language models' performance and energy use, with strategies to optimize latency and efficiency.
Reframing AI-Supported Writing as Creative Blackout Poetry
Explore how human-AI interaction traces can be seen as blackout poetry, enhancing creativity and trust in AI-assisted writing processes.
ECHO: Fast Speculative Decoding for High-Concurrency LLMs
ECHO boosts large language model inference with elastic speculative decoding and sparse gating, achieving up to 5.35x speedup in high-concurrency scenarios...
ML-Based Duration Scheduler Cuts HPC Job Wait Times
Optimize HPC workload scheduling with ML-driven job duration predictions, reducing wait times by 11% and boosting system throughput.
Executable Frameworks for Creative Emergence & Foresight
Discover innovative executable protocols that turn creativity theories into practical frameworks for strategic foresight and creative emergence.