Fix issue 394 and update info#481

Merged

ooples merged 277 commits intomasterfrom

claude/fix-issue-394-011CV3HkgfwwbaSAdrzrKd58

Nov 17, 2025

This pull request is big! We're only showing the most recent 250 commits

Commits on Nov 13, 2025

Commits on Nov 14, 2025

Commits on Nov 15, 2025

Commits on Nov 16, 2025

fix: expectedsarsa modelmetadata now includes featurecount and complexity
ooples
committed
fix: qmix agent serialization and save/load now properly implemented
ooples
committed
fix: prevent negative infinity in thompson sampling beta distribution

ooples
and
claude
committed
fix: tabular qlearning setparameters now preserves state keys
ooples
committed
fix: resolve 10 P0 critical issues for agent 5 work package

ooples
and
claude
committed
fix: correct weighted importance sampling and watkins q lambda trace reset

ooples
and
claude
committed
fix: add maddpg validation and prevent a2c null options crash in base initializer
ooples
committed
fix(dreamer): correct batch processing and gradient calculations

ooples
and
claude
committed
style(dreamer): remove fix comments from code
ooples
committed
fix(maddpg): use target actors for target Q computation

ooples
and
claude
committed
fix(qmix): enforce monotonicity in mixing network weights

ooples
and
claude
committed
fix(dqn): sync target network in setparameters

ooples
and
claude
committed
fix(sarsa): setparameters cannot restore Q-values without state keys

ooples
and
claude
committed
fix(doubleqlearning): setparameters cannot restore without state information

ooples
and
claude
committed
fix(decisiontransformer): use optimizer for gradient application

ooples
and
claude
committed
fix(nstepqlearning): setparameters cannot restore without state information
ooples
committed
style(epsilongreedybanditoptions): remove unused using directives
ooples
committed
style(sarsalambdaoptions): remove unused using directives
ooples
committed
fix(dynaq): applygradients throws exception for unsupported operation
ooples
committed
fix(dreamer): applygradients throws exception for multi-network complexity
ooples
committed
fix(montecarlo): savemodel/loadmodel throw exception for unsupported persistence
ooples
committed
fix: remove resetgradients call from ddpg agent
ooples
committed
fix: correct ppoagent cliprange property and linearsarsaagent deserialize null handling
ooples
committed
fix: add comprehensive null checks for td3agent discountfactor and minqtarget
ooples
committed
chore: remove investigation/report files and temporary scripts per CLAUDE.md guidelines
ooples
committed
Delete POLICY_BASE_CLASSES_PROMPT.md
ooples
authored
Delete src/ReinforcementLearning/INTEGRATION_PLAN.md
ooples
authored
Delete fix-ppo-rainbow-dueling-muzero.sh
ooples
authored
fix: resolve merge conflicts with master
ooples
committed
fix: add savestate and loadstate methods to reinforcement learning agent base

ooples
and
claude
committed
Merge branch 'claude/fix-issue-394-011CV3HkgfwwbaSAdrzrKd58' of https://github.com/ooples/AiDotNet into claude/fix-issue-394-011CV3HkgfwwbaSAdrzrKd58
ooples
committed
fix: add missing using directive for jsonconvert in lspi agent
ooples
committed
fix: add validation for state and action size in expected sarsa options
ooples
committed
fix: add defensive validation for expectedsarsaagent options

ooples
and
claude
committed
fix: add constructor validation to expectedsarsaoptions
ooples
committed
fix: add deployment configuration to rl training path
ooples
committed
fix: replace dummy gradient vector with notsupportedexception in duelingdqnagent

ooples
and
claude
committed

Commits on Nov 17, 2025