Clarify two-checkpoint roles: memory-manager vs answer-agent

README.md

@@ -10,28 +10,29 @@ tags:
   - agent
 ---
 
-# Memory-R2 7B — Memory Manager (LoGo-GRPO, 32-session champion)
+# Memory-R2 7B — Memory Manager + Answer Agent
 
-This is the trained **memory-management policy** from [Memory-R2: Fair Credit Assignment for Long-Horizon Memory-Augmented LLM Agents](https://arxiv.org/abs/2605.21768) (arXiv:2605.21768).
+Checkpoints from [Memory-R2: Fair Credit Assignment for Long-Horizon Memory-Augmented LLM Agents](https://arxiv.org/abs/2605.21768) (arXiv:2605.21768).
 
-It is a Qwen2.5-7B-Instruct model fine-tuned with **LoGo-GRPO** (turn-level + token-level credit assignment) via a curriculum of 8 → 16 → 32-session rollouts on the LoCoMo long-horizon dialogue dataset. Given a running conversation, it decides what to INSERT / UPDATE / DELETE in an external memory store, so that a separate answer agent can later answer questions using only the maintained memory.
+This repo contains **two separate Qwen2.5-7B-Instruct checkpoints** that play two different, non-overlapping roles in the pipeline. They are not interchangeable and neither one does the other's job:
 
-This checkpoint is the paper's deployed "champion" (`32sess_champion_v2`, LoGo-GRPO curriculum, global step 5) — the memory manager behind the headline `tab:main` **Memory-R2 (OSS)** result on LoCoMo:
+| Subfolder | Role | Does it manage memory? | Does it answer questions? |
+| --- | --- | :---: | :---: |
+| **`memory-manager/`** | Stage 1 — reads the running conversation and decides what to INSERT / UPDATE / DELETE in the external memory store. | ✅ Yes — this is its only job | ❌ No — it never sees or answers the held-out QA questions |
+| **`answer-agent/`** | Stage 2 — given a question and the memory store `memory-manager/` produced, generates the final answer. | ❌ No — it never touches the memory store's write operations | ✅ Yes — this is its only job |
 
-| Answer agent | F1 | BLEU-1 | LLM-judge (gpt-4o-mini) |
-| --- | ---: | ---: | ---: |
-| GPT-OSS-120B (untrained, paired at eval time) | **49.29** | **43.64** | **86.08** |
+**`memory-manager/`** is `32sess_champion_v2` — the LoGo-GRPO-trained (turn-level + token-level credit assignment, curriculum 8→16→32 sessions on LoCoMo) memory-management policy, and is the paper's main contribution / "the champion" checkpoint. It is the piece you need regardless of which answer agent you pair it with.
 
-See the paper's `tab:different-answer-agent` for how this same memory manager performs when paired with other answer agents (untrained Qwen-7B, RL-trained Qwen-7B, etc.).
+**`answer-agent/`** (`sft_cont_step55`, SFT+RL-trained) is an *optional, swappable* stage-2 model — the memory manager was evaluated in the paper against several different answer agents, and this is simply the best-performing one we trained ourselves. You can equally pair `memory-manager/` with an untrained Qwen-7B, GPT-OSS-120B, or any other instruction-tuned LLM as the answer agent — see `tab:different-answer-agent` in the paper. The reverse never happens: `answer-agent/` is not used for memory operations, and swapping it out has no effect on how memory is maintained.
 
-## Answer agent (`answer-agent/`)
+## Headline results (`tab:main`)
 
-Also included is `sft_cont_step55`, a separately SFT+RL-trained Qwen2.5-7B-Instruct **answer agent**: given the memory manager's maintained memory store, it answers the held-out questions. Paired with the `memory-manager/` checkpoint above, this is the paper's highest-F1 configuration ("Memory-R2 (SFT-RL)" in `tab:main`):
+`memory-manager/` is the constant in every row below; only the answer agent changes:
 
 | Memory manager | Answer agent | F1 | BLEU-1 | LLM-judge (gpt-4o-mini) |
 | --- | --- | ---: | ---: | ---: |
-| memory-manager/ | answer-agent/ | **51.46** | **44.84** | 69.03 |
-| memory-manager/ | GPT-OSS-120B (untrained) | 49.29 | 43.64 | **86.08** |
+| memory-manager/ | `answer-agent/` (ours, SFT+RL) | **51.46** | **44.84** | 69.03 |
+| memory-manager/ | GPT-OSS-120B (untrained, external) | 49.29 | 43.64 | **86.08** |
 
 ## Usage