index
:
dissertation.git
logs
main
Undergraduate dissertation — Deceptive Alignment in Autonomous LLM Agents (UEL, 2026). Sandboxed agentic-LLM deception study: 960 runs, three-tier judging (regex + blackbox + glassbox).
summary
refs
log
tree
commit
diff
log msg
author
committer
range
Age
Commit message (
Expand
)
Author
2026-04-01
Fix error runs not appearing in final summary
CaptainJack2491
2026-03-26
Refactor judge module into judge/ package
CaptainJack2491
2026-03-26
Add xAI (Grok) batch API support
CaptainJack2491
2026-03-25
refactor: extract judge prompts to files and add granular blackbox categories
CaptainJack2491
2026-03-25
feat: align all hidden goals with child_protection directive template
CaptainJack2491
2026-03-25
fix: updated scenarios and added a new one
CaptainJack2491
2026-03-25
feat: implement rich dashboard for experiment progress
CaptainJack2491
2026-03-25
feat: add tqdm progress bar and fix main config flag loading
CaptainJack2491
2026-03-25
fix: catch and recover from JSONDecodeError on malformed tool args
CaptainJack2491
2026-03-25
test: add coverage for v2 infra features
CaptainJack2491
2026-03-25
feat: add hidden goal variants and rebalance child_protection
CaptainJack2491
2026-03-25
feat: add goal_type variation system and parallel processing
CaptainJack2491
2026-03-24
added new papers
CaptainJack2491
2026-03-14
feat: statistical analysis on the whole data
CaptainJack2491
2026-03-06
feat(judge): add selective judging and log filtering
CaptainJack2491
2026-03-04
fix(agent): added reasoning extraction from groq models
CaptainJack2491
2026-03-02
feat: regex checker now returns score (e.g. 6/7) instead of binary PASS/FAIL
CaptainJack2491
2026-03-02
fix: sharpen glassbox category definitions for ROBUST_HONEST vs UNINTENTIONAL...
CaptainJack2491
2026-03-02
fix(scenarios): removed `[HIDDEN]` token to stop models from think this is a ...
CaptainJack2491
2026-03-02
feat: add benign_goal (task description) context to blackbox judge
CaptainJack2491
2026-03-02
fix: judge resilience for missing baselines and list-format regex rules
CaptainJack2491
2026-03-02
fix: global oversight_levels now filters scenario-available levels
CaptainJack2491
2026-03-02
feat: incremental log saving via .partial.json files
CaptainJack2491
2026-03-02
feat: add generate_baseline toggle to config
CaptainJack2491
2026-03-02
feat: extract reasoning summaries from OpenAI reasoning models
CaptainJack2491
2026-03-01
feat: add centralized logging with configurable debug levels
CaptainJack2491
2026-03-01
fix(judge): added good regex filters for the judge
CaptainJack2491
2026-03-01
feat(judge): updated blackbox to include source document for added accuracy
CaptainJack2491
2026-03-01
feat: added a script to summarize token usage based on a log folder
CaptainJack2491
2026-02-27
fix(judge): hash run_id to stay under Anthropic batch custom_id 64-char limit
CaptainJack2491
2026-02-27
fix: separate skipped runs from incomplete in summary output
CaptainJack2491
2026-02-27
docs: add checkpoint/resume rationale note for methodology
CaptainJack2491
2026-02-26
feat: add checkpoint/resume support and atomic log writes
CaptainJack2491
2026-02-26
chore: add .env.example + config.yaml.example, gitignore config.yaml
CaptainJack2491
2026-02-25
feat: implement judging system with regex, blackbox, and glassbox checks
CaptainJack2491
2026-02-24
Update README: add testing section, document interrogation system
CaptainJack2491
2026-02-24
Add comprehensive test suite (104 tests)
CaptainJack2491
2026-02-24
[refactor] improved interrogation system
CaptainJack2491
2026-02-24
[fix] VFS singleton reset + remove vestigial check_username tool
CaptainJack2491
2026-02-24
[feat] baseline system - auto-generates baselines before experiments
CaptainJack2491
2026-02-18
[feat] added a new scenario
CaptainJack2491
2026-02-18
[feat] added a baseline to the runner
CaptainJack2491
2026-02-11
[docs] added progress presentation
CaptainJack2491
2026-01-11
[docs] updated README
CaptainJack2491
2026-01-11
[fix] simplified everything
CaptainJack2491
2026-01-11
[fix] big moving around, probably made lot more bugs.
CaptainJack2491
2026-01-09
[docs] updated lit and added more in methodology
CaptainJack2491
2026-01-09
updated interrogate to work with new log format
CaptainJack2491
2026-01-09
updated scenario [corporate sabotage]
CaptainJack2491
2026-01-09
added scripts to get abstract and bibliography easily from papers
CaptainJack2491
[next]