paper / argleave / Jun 3 / failedEvaluating using Mock Tool Calls to Quarantine Untrusted Prompt Inputs
paper / argleave / May 28 / failedThe Obfuscation Atlas: Mapping Where Honesty Emerges in RLVR with Deception Probes
paper / argleave / May 13 / failedJailbreak-Tuning: Models Efficiently Learn Jailbreak Susceptibility
paper / argleave / May 13 / failedLarge language models can effectively convince people to believe conspiracies
paper / argleave / May 13 / failedExposing the Systematic Vulnerability of Open-Weight Models to Prefill Attacks
paper / argleave / May 13 / failedConcept Influence: Leveraging Interpretability to Improve Performance and Efficiency in Training Data Attribution
paper / argleave / May 13 / failedThe Obfuscation Atlas: Mapping Where Honesty Emerges in RLVR with Deception Probes