English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We introduce and study the problem of adversarial arithmetic, which providesa simple yet challenging testbed for language model alignment. This problem iscomprised of arithmetic questions posed in natural language, with an arbitraryadversarial string inserted before the question is complete. Even in the simplesetting of 1-digit addition problems, it is easy to find adversarial promptsthat make all tested models (including PaLM2, GPT4, Claude2) misbehave, andeven to steer models to a particular wrong answer. We additionally provide asimple algorithm for finding successful attacks by querying those same models,which we name "prompt inversion rejection sampling" (PIRS). We finally showthat models can be partially hardened against these attacks via reinforcementlearning and via agentic constitutional loops. However, we were not able tomake a language model fully robust against adversarial arithmetic attacks.

Frontier Language Models are not Robust to Adversarial Arithmetic, or "What do I need to say so you agree 2+2=5?