added ba_two_motifs subset, re-generated all subsets

.gitignore

@@ -5,4 +5,5 @@
 __pycache__/
 
 # Debugging
-fig/*/*.pdf
+fig/*/pos_graphs/*.pdf
+fig/*/neg_graphs/*.pdf

README.md

@@ -42,21 +42,31 @@ configs:
         path: data/tree_grid/val.jsonl
       - split: test
         path: data/tree_grid/test.jsonl
+  - config_name: ba_two_motifs
+    data_files:
+      - split: train
+        path: data/ba_two_motifs/train.jsonl
+      - split: validation
+        path: data/ba_two_motifs/val.jsonl
+      - split: test
+        path: data/ba_two_motifs/test.jsonl
 ---
 
 # MotifQA
 
 ## Dataset Summary
 
-MotifQA is a synthetic graph question-answering benchmark focused on detecting the presence of a cetain motif inside small random graphs.
+MotifQA is a synthetic graph question-answering benchmark focused on detecting graph motifs inside small random graphs.
 Each example pairs a textual prompt with an answer sentence, a list of nodes highlighted as the motif (when present), and an explicit graph description(nodes and edges).
 In this QA dataset, all graphs are homogenous and undirected.
+Subsets cover both yes/no motif detection and motif-type classification (house vs 5-cycle).
 
 ## Subsets
 
 - `ba_shapes` (default): Detect a five-node house motif inside Barabási–Albert graphs.
 - `tree_cycle`: Detect a 6-node cycle inside random binary trees.
 - `tree_grid`: Detect a 3×3 grid motif inside random binary trees.
+- `ba_two_motifs`: Classify whether a Barabási–Albert graph contains a house motif or a 5-cycle motif (every graph has exactly one of the two).
 
 ## Dataset Structure
 
@@ -74,17 +84,23 @@ In this QA dataset, all graphs are homogenous and undirected.
 
 | Subset     | Split | Total | Positive | Negative |
 | ---------- | ----- | ----- | -------- | -------- |
-| ba_shapes  | Train | 1000  | 492      | 508      |
-| ba_shapes  | Val   | 500   | 255      | 245      |
-| ba_shapes  | Test  | 500   | 253      | 247      |
-| tree_cycle | Train | 1000  | 497      | 503      |
-| tree_cycle | Val   | 500   | 246      | 254      |
-| tree_cycle | Test  | 500   | 257      | 243      |
-| tree_grid  | Train | 1000  | 489      | 511      |
-| tree_grid  | Val   | 500   | 268      | 232      |
-| tree_grid  | Test  | 500   | 243      | 257      |
-
-Positive and negative graphs are approximately balanced across splits.
+| ba_shapes  | Train | 1000  | 511      | 489      |
+| ba_shapes  | Val   | 500   | 246      | 254      |
+| ba_shapes  | Test  | 500   | 243      | 257      |
+| tree_cycle | Train | 1000  | 496      | 504      |
+| tree_cycle | Val   | 500   | 245      | 255      |
+| tree_cycle | Test  | 500   | 259      | 241      |
+| tree_grid  | Train | 1000  | 508      | 492      |
+| tree_grid  | Val   | 500   | 242      | 258      |
+| tree_grid  | Test  | 500   | 250      | 250      |
+
+Detection subsets are approximately balanced across splits. The `ba_two_motifs` subset contains only motif-bearing graphs; label counts are:
+
+| Subset        | Split | Total | House | Cycle |
+| ------------- | ----- | ----- | ----- | ----- |
+| ba_two_motifs | Train | 1000  | 513   | 487   |
+| ba_two_motifs | Val   | 500   | 241   | 259   |
+| ba_two_motifs | Test  | 500   | 246   | 254   |
 
 ### Node Count Distribution
 
@@ -96,6 +112,8 @@ Node counts depend on the underlying base generator and motif size:
   ![Tree cycle node counts](fig/tree_cycle/node_count_hist.svg)
 - `tree_grid` (Prüfer trees + 3×3 grid motif): 8–24 nodes.
   ![Tree grid node counts](fig/tree_grid/node_count_hist.svg)
+- `ba_two_motifs` (Barabási–Albert; house vs 5-cycle motifs): 8–20 nodes.
+  ![BA two motifs node counts](fig/ba_two_motifs/node_count_hist.svg)
 
 Observed frequencies across the full 2,000-example splits for each subset are shown in the per-subset histograms.
 
@@ -105,12 +123,12 @@ Observed frequencies across the full 2,000-example splits for each subset are sh
 
 All examples are synthetically generated using [PyTorch Geometric](https://pytorch-geometric.readthedocs.io/) (2.7.0) primitives:
 
-- Positive samples start from certain base graphs whose node counts are drawn from a uniform distribution before a motif is injected, yielding final graphs with 8–20 nodes.
-- Negative samples use the same process with node counts drawn from `range(MIN_NODES, MAX_NODES + 1)` (default 8–20) and no motif.
+- Detection subsets: positive samples start from certain base graphs whose node counts are drawn from a uniform distribution before a motif is injected, yielding final graphs with 8–20 nodes. Negative samples use the same process with node counts drawn from `range(MIN_NODES, MAX_NODES + 1)` (default 8–20) and no motif.
+- `ba_two_motifs`: both classes contain motifs injected into Barabási–Albert graphs—house motifs vs 5-cycle motifs—with final graphs spanning 8–20 nodes.
 
 ### Generation Process
 
-1. Create positive (`ExplainerDataset`) and negative (`NegativeGraphs`) graph pools.
+1. Create class-specific graph pools (`ExplainerDataset` for motif-bearing graphs and `NegativeGraphs` for motif-free graphs, or two different `ExplainerDataset` instances for `ba_two_motifs`).
 2. Randomly shuffle node indices via the custom `ShuffleNodes` transform to prevent positional bias.
 3. Convert graphs to text QA format:
    - Construct the prompt and answer sentence.