# AI Research Tracker — batch ingest sources
# Run: synthadoc ingest --batch sources.txt
#
# These are the sources used in the Synthadoc demo walkthrough.
# Lines starting with # are comments and are ignored.

# Public-domain source files (synced via: synthadoc demo sync ai-research)
raw_sources/public-domain/transformer-architecture.txt
raw_sources/public-domain/attention-mechanisms.txt
raw_sources/public-domain/large-language-models.txt
raw_sources/public-domain/reinforcement-learning-from-human-feedback.txt
raw_sources/public-domain/scaling-laws.txt
raw_sources/public-domain/training-techniques.txt
raw_sources/public-domain/geoffrey-hinton-biography.txt
raw_sources/public-domain/andrej-karpathy-biography.txt

# Local file: benchmark methodology correction (contradicts llm-benchmarks.md)
raw_sources/llm-benchmarks-q1-2026.pdf

# Web and URL sources (demonstrate URL ingestion and search intent)
https://arxiv.org/abs/2212.08073
https://www.youtube.com/watch?v=kCc8FmEb1nY
search for: chain-of-thought prompting research 2024
search for: speculative decoding inference optimization
search for: mixture of experts LLM architecture 2024
