Tootfinder

Prompt Engineering: Tipps für den Einsatz generativer KI im Unternehmen
Im Webinar am 14. Mai gibt der Fachdienst heise KI PRO einen Überblick über das Thema Prompt Engineering und die wichtigsten Inhalte des aktuellen Deep Dives.

KI-Update kompakt: Geld verdienen, OpenAI API, Prompt-Safety, Newsletter
Das "KI-Update" liefert werktäglich eine Zusammenfassung der wichtigsten KI-Entwicklungen.
https://www.

This https://arxiv.org/abs/2311.04934 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

Prompt Cache: Modular Attention Reuse for Low-Latency Inference
We present Prompt Cache, an approach for accelerating inference for large language models (LLM) by reusing attention states across different LLM prompts. Many input prompts have overlapping text segments, such as system messages, prompt templates, and documents provided for context. Our key insight is that by precomputing and storing the attention states of these frequently occurring text segments on the inference server, we can efficiently reuse them when these segments appear in user prompts.…

Oh fuck no. Logitech, don't give me this bullshit. Get AI out of my mouse!
https://www.theverge.com/2024/4/17/24132468/logitech-ai-prompt-builder-button

Logitech wants you to press its new AI button
Logitech wants to encourage more people to use AI with its new Logi AI Prompt Builder, which makes it easier to use ChatGPT. It will also release a new mouse with an AI button.

Investigating the prompt leakage effect and black-box defenses for multi-turn LLM interactions
Divyansh Agarwal, Alexander R. Fabbri, Philippe Laban, Shafiq Joty, Caiming Xiong, Chien-Sheng Wu
https://arxiv.org/abs/2404.16251

Investigating the prompt leakage effect and black-box defenses for multi-turn LLM interactions
Prompt leakage in large language models (LLMs) poses a significant security and privacy threat, particularly in retrieval-augmented generation (RAG) systems. However, leakage in multi-turn LLM interactions along with mitigation strategies has not been studied in a standardized manner. This paper investigates LLM vulnerabilities against prompt leakage across 4 diverse domains and 10 closed- and open-source LLMs. Our unique multi-turn threat model leverages the LLM's sycophancy effect and our ana…

Auch Neues von Deerhof (bzw. Greg Saunier). #sound
https://

Congress considers legislation to block DJI drones from running on US communication networks, effectively killing the Chinese company's US commercial business (Kate Kelly/New York Times)
https://www.nytimes.com/2024/04/…

A Chinese Firm Is America’s Favorite Drone Maker. Except in Washington.
U.S. authorities consider DJI a security threat. Congress is weighing legislation to ban it, prompting a lobbying campaign from the company, which dominates the commercial and consumer drone markets.

Neste papo sobre #Babylon5, olha o que eu acabei achando pra aluguel na amazon prime: https://www.primevideo.com/detail/0PI0YDKAZRVTFIX7TPG5A1MJ…

When and how will US aid impact the front lines in Ukraine?: https://benborges.xyz/2024/04/25/when-and-how.html

Auch Neues von Deerhof (bzw. Greg Saunier). #sound
https://