Extending Activation Steering to Broad Skills and Multiple Behaviours von van der Weij, Teun, Poesio, Massimo, Schoots, Nandi

Volltext bestellen

Evaluating Shutdown Avoidance of Language Models in Textual Scenarios von van der Weij, Teun, Lermen, Simon, lang, Leon

Volltext bestellen

AI Sandbagging: Language Models can Strategically Underperform on Evaluations von van der Weij, Teun, Hofstätter, Felix, Jaffe, Ollie, Brown, Samuel F, Ward, Francis Rhys

Volltext bestellen

Noise Injection Reveals Hidden Capabilities of Sandbagging Language Models von Tice, Cameron, Kreer, Philipp Alexander, Helm-Burger, Nathan, Shahani, Prithviraj Singh, Ryzhenkov, Fedor, Haimes, Jacob, Hofstätter, Felix, van der Weij, Teun

Volltext bestellen

Extending Activation Steering to Broad Skills and Multiple Behaviours von Teun van der Weij, Poesio, Massimo, Nandi Schoots

Volltext

Evaluating Shutdown Avoidance of Language Models in Textual Scenarios von Teun van der Weij, Lermen, Simon, lang, Leon

Volltext

AI Sandbagging: Language Models can Strategically Underperform on Evaluations von Teun van der Weij, Hofstätter, Felix, Jaffe, Ollie, Brown, Samuel F, Ward, Francis Rhys

Volltext

Noise Injection Reveals Hidden Capabilities of Sandbagging Language Models von Tice, Cameron, Kreer, Philipp Alexander, Helm-Burger, Nathan, Prithviraj Singh Shahani, Ryzhenkov, Fedor, Haimes, Jacob, Hofstätter, Felix, Teun van der Weij

Volltext