RSSAmplifier

Blog

Sinatras

Sinatras RSS

blog.sinatras.devRSS feed ↗17 posts

Latest posts

PMPP-Eval Journey

Investigating Inaction Biases in Models Through World Simulation

Carla-Env

KernelBot-Hacking

KernelGuard

Reward Hacking in KernelBench

Audience Toggle Variants Test

Pushing the Boundaries of Realism

research-env General Agent Experiment

SynthRewardHacking

Hedonic Treadmill of Proxy Reward Hacking

PMPP-Hard Correct Is Not Fast Enough

PMPP-Hard-Blog

PMPP-Hard-Blog-reordered-draft

PMPP-Hard-Blog-reordered-german-pipeline-draft

PMPP-Hard-Blog-reordered-german-pipeline-draft 1

PMPP-Hard