Spaces:
Sleeping
Sleeping
Commit History
-
Merge pull request #2 from KrisKeshav/cleanup-dead-code
KrisKeshavUnverified signatureRemove dead code left over from holdover-controller iterations28040d1 -
Remove dead code left over from holdover-controller iterations
- environment: unused θ̂ history buffer and _theta_hat_at helper - sensors: unused PMUStream.arrived(), unused engage() argument - pll: unused SRFPLL.snapshot() - unused imports (pyflakes clean) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>ff68477 -
Merge pull request #1 from KrisKeshav/v2-physics-overhaul
KrisKeshavUnverified signaturev2: physically correct SRF-PLL, realistic attacks, exploit-proof graders A completely improved version of the earlier work7e03bda -
v2: physically correct SRF-PLL, realistic attacks, exploit-proof graders
Physics - Fix Clarke/Park transforms (vq was cos(θ+θ̂), so the PLL never locked); cosine-referenced amplitude-invariant transforms, all three phases, 5 kHz simulation, ωn=100 rad/s, ζ=0.707, amplitude-normalised PI loop - Natural grid events (phase jumps, sags, single-phase faults, harmonic steps, frequency events) as confounders; sensor noise + 12-bit ADC - Coordinated phase-spoofing attack for the hard task; trusted, delayed PMU stream; PMU-disciplined holdover and latched disconnect actions - Desync measured against the true grid angle; shadow PLL gives the counterfactual time-to-desync used by the hard grader Environment / graders / reward - 40 decisions x 50 ms, attack-free episodes, single seeded RNG tree - Ground truth hidden until done; env detector no longer leaked in info - Graders penalise false alarms; always/never/random policies score low, oracle scores >= 0.9 on every seed - Dense reward with Brier calibration, mitigation costs and harm penalty OpenEnv / in…c282634
-
Update README.md
KrisKeshavUnverified signatureec5384c
-
Refactor: Restore intrinsic detector to fallback logic, rewrite README.md, and polish all codebase comments for final submission
01f8cd5 -
Add LLM frame skip every 10 steps + clamp all scores to strict (0,1)
a8f21d0
-
Clamp grader scores to strict (0,1) range for validator compliance
8a428a8 -
Restore 500 steps, LLM every step with circuit breaker, use requests.Session for speed
4c2a495 -
Decrease MAX_STEPS to 250 and change LLM frame skip to 10
82e7138 -
Optimize LLM frame rate to fix evaluation timeout
14e5865 -
Fix Dockerfile curl Healthcheck and obey stdout protocols
56a4099 -
Fix inference.py timeout crashes and unhandled exceptions
eda1886 -
fix(inference): align OpenEnv strict proxy compliance overrides and remove hardcoded API validations
be25efb -
chore: add bash validator and format asyncio logic
e897efa -
fix(inference): refactor proxy initialization and baseline logic
4b406c3
-
fix(inference): final cleanup and user modifications
cd2d585 -
fix(inference): harden execution safety and proxy compliance
a437a7c -
fix: add Hugging Face Spaces configuration to README
03f87aa -
ci: remove force push from deployment workflow
a887b44 -
ci: trigger initial sync with force push
492cfef -
ci: Add GitHub Actions workflow for HF Space deployment
1510c68 -
Update: Sync all modules, add detector/tests/validation, fix inference agent logic
20bc5e4
-
Refine comments and environment variable details
KrisKeshavUnverified signatureUpdated comments and environment variable descriptions for clarity.153732e -
Refactor pll_sim.py for code clarity
KrisKeshavUnverified signatureRemoved unnecessary blank lines and improved comment clarity.b8a96e0 -
small changes
KrisKeshavUnverified signatureRemove redundant comments in coerce_attack_params method.1aaa98d -
Improving comments in grading functions for better understanding
KrisKeshavUnverified signatureRefactor comments for clarity in grading functions.c53a16c -
Fix PLL simulation issues and improve observation handling
KrisKeshavUnverified signatureThis commit includes several fixes and improvements to the PLL simulation environment, including updates to the handling of attack signals, observation windows, and episode termination conditions.3f7ca79 -
small changes
KrisKeshavUnverified signaturef4bc318 -
Update docstring for server entry point
KrisKeshavUnverified signaturec54ccdc -
Enhance API endpoints with async locking and error handling
KrisKeshavUnverified signatureRefactor /reset, /step, and /state endpoints to use async locks for thread safety. Update ResetRequest to handle optional seed.542893e -
Initial commit: PLL Cyberattack Detection OpenEnv
81e328b