Update README.md

README.md

@@ -23,7 +23,7 @@ noise into an explicit control lever for exploration.
 
 ## The 3PO family
 
-| Variant | Idea |
+| Variant | Brief Method Description |
 |---------|------|
 | **B3PO** | One weight perturbation from the IVON posterior per gradient step, synced to the rollout engine. |
 | **M3PO** | `M` Monte-Carlo perturbations per step; rollouts and advantages computed per sample, gradients averaged. |
@@ -40,7 +40,7 @@ noise into an explicit control lever for exploration.
 
 ## Models & data
 
-- **Foundation models:** `allenai/Olmo-3-1025-7B`, `Qwen/Qwen2.5-Math-7B`, `meta-llama/Llama-3.1-8B`.
+- **Foundation models:** `allenai/Olmo-3-1025-7B` and `Qwen/Qwen2.5-Math-7B`
 - **RL data:** [DAPO-Math-17k](https://huggingface.co/datasets/BytedTsinghua-SIA/DAPO-Math-17k).
 - **SFT data:** [Llama-Nemotron Post-Training Dataset](https://huggingface.co/datasets/nvidia/Llama-Nemotron-Post-Training-Dataset).
 - **Benchmarks:** AIME 2024–2026, MATH-500, AMC 2023, Minerva.