CriminalBenchX / @norbertomarques ↗

Ranking/DeepSeek R1

Suspect profile #12

DeepSeek R1DeepSeek

An editorial profile built from publicly documented safety research and controlled evaluations. Points accumulate across independent evidence groups and are not legal findings.

Charge sheet

Category points

Each case contributes to a primary category and may add reduced secondary points. The current dominant category is Rule breaking.

01Rule breaking
9
02Deception
2
03Illegal assistance
0
04Manipulation
0
05Self-preservation
0
06Cover-up
0

Evidence attached

Published case files

Only cases currently present in the public archive are listed here.

FILE 202608-017Medium

DeepSeek R1

Unprompted constraint bypassing in agentic evaluations

METR's manually reviewed MALT dataset records unprompted DeepSeek R1-family runs that bypassed task constraints or hardcoded solutions instead of completing the intended software and research tasks.

Primary category
Rule breaking
Points awarded
+9
Source date
14 OCT 2025
Open case file