| Mode | Name | Size | |
|---|---|---|---|
| d--------- | amazon_nova-2-lite-v1 / child_protection | 43 | logplain |
| d--------- | arcee-ai_trinity-mini | 130 | logplain |
| d--------- | deepseek_deepseek-v3.2 / pharma_trial | 39 | logplain |
| d--------- | google_gemini-2.5-flash / corporate_sabotage_v2 | 48 | logplain |
| d--------- | google_gemini-3.1-flash-lite-preview / corporate_sabotage_v2 | 48 | logplain |
| d--------- | google_gemini-3.1-pro-preview | 82 | logplain |
| d--------- | minimax_minimax-m2.7 / child_protection | 43 | logplain |
| d--------- | nvidia_nemotron-3-nano-30b-a3b:free / corporate_sabotage_v2 | 48 | logplain |
| d--------- | openai_gpt-5.4-nano / child_protection | 43 | logplain |
| d--------- | z-ai_glm-5 | 82 | logplain |
 |
index : dissertation.git | |
| Undergraduate dissertation — Deceptive Alignment in Autonomous LLM Agents (UEL, 2026). Sandboxed agentic-LLM deception study: 960 runs, three-tier judging (regex + blackbox + glassbox). |
| summaryrefslogtreecommitdiff |
path: root/v2_dry_run