Study evaluates DeepSeek V4.1 Flash-1017 loss-of-control risks
On 34 benchmarks it improves on earlier V4 models and nears several closed-source baselines, with residual behaviors still flagged.
1 stories, newest first · one card per story
On 34 benchmarks it improves on earlier V4 models and nears several closed-source baselines, with residual behaviors still flagged.
Today 692 readers · 859 views · 653 reads · all time 159K views