Add phoneme-to-word matching tests for BCI dataset

- Extracted 1718 phoneme-word pairs from t15_copyTask.pkl
- Implemented 3 testing strategies for ZeroEntropy
- Tested on full dataset: 40% success rate with 1718 docs, 100% with 100 docs
- Created comprehensive documentation and integration notes
- Added Python extraction script for pickle data
- All emojis removed from documentation per request
- Results show ZeroEntropy viable for BCI augmentation with hybrid approach

Author: davidatoms

.gitignore

@@ -30,3 +30,6 @@ GUIDE.md
 # Claude AI assistant directory
 .claude/
 docs/
+
+# Future integration notes (internal planning)
+future-integrations/

CHANGELOG.md

@@ -23,3 +23,24 @@ tokio-test = "0.4"
 dotenv = "0.15"
 csv = "1.3"
 chrono = "0.4"
+clap = { version = "4.5", features = ["derive"] }
+
+[[example]]
+name = "search_gpt2_dataset"
+path = "examples/search_gpt2_dataset.rs"
+
+[[example]]
+name = "emergence_tester"
+path = "examples/emergence_tester.rs"
+
+[[example]]
+name = "phoneme_to_word_bci"
+path = "examples/phoneme_to_word_bci.rs"
+
+[[example]]
+name = "phoneme_to_word_advanced"
+path = "examples/phoneme_to_word_advanced.rs"
+
+[[example]]
+name = "phoneme_to_word_full_dataset"
+path = "examples/phoneme_to_word_full_dataset.rs"