2 papers
cs.CL2026
DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
DeepSeek-AI, :, Anyi Xu +585
The widespread adoption of long-horizon agents has made model workloads increasingly input-heavy. Although prior work has substantially reduced the cost of long-context computation…
cs.CL2022
ConsPrompt: Exploiting Contrastive Samples for Fewshot Prompt Learning
Jinta Weng, Yifan Deng, d Donghao Li +3
The prompt has become an effective linguistic tool for utilizing pre-trained language models. However, in few-shot scenarios, subtle changes in the prompt design always make the re…