Dataset Viewer

The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.

Chimere Engram Tables

Pre-built N-gram hash tables for the Chimere multi-tier Engram memory system. These provide O(1) lookup for domain-specific token sequences, used to bias LLM generation with factual knowledge at the logit level.

Files

File Size Domain N-grams
code.engr 155 KB Programming / software Code patterns, API signatures
cyber.engr 172 KB Cybersecurity Threat indicators, CVE patterns
general.engr 679 KB General knowledge Common factual patterns

Format

Binary hash tables with FNV-1a hashing, compatible with the Rust EngramLookup implementation in chimere-server and the Python tools in chimere-odo.

Structure

  • 20-byte header (magic, version, n-gram count, table size)
  • Hash table slots with FNV-1a keys
  • O(1) lookup, <10ns with Cuckoo filter pre-screening

Usage

# Query an Engram table
python engram_query.py code.engr "async def main"

# Build your own from documents
python engram_ingest.py --input docs/ --output custom.engr

Author

Kevin Remondiere — Independent ML researcher

Downloads last month
36