CivMarshmallow Checkpoints for Rome Wasn't Built in One Turn: Replay-Grounded Reward Signals for Long-Horizon Freeciv Agents CivMarshmallow/CivTelescope Reinforcement Learning • Updated 19 days ago CivMarshmallow/CivMarshmallow-Qwen3-8B Reinforcement Learning • Updated 6 days ago CivMarshmallow/Baselines-Qwen3-8B Reinforcement Learning • Updated 6 days ago
CivMarshmallow Checkpoints for Rome Wasn't Built in One Turn: Replay-Grounded Reward Signals for Long-Horizon Freeciv Agents CivMarshmallow/CivTelescope Reinforcement Learning • Updated 19 days ago CivMarshmallow/CivMarshmallow-Qwen3-8B Reinforcement Learning • Updated 6 days ago CivMarshmallow/Baselines-Qwen3-8B Reinforcement Learning • Updated 6 days ago