reaperdoesntknow/Symiotic-14B

TEXT GENERATIONPricing:Input $0.48 / Output $0.96Concurrent Unit Cost:1Model Size:14BQuant:FP8Context Size:32kTool Calling:SupportedPublished:May 6, 2025License:afl-3.0Architecture:Transformer0.0K Featherless Exclusive Cold

reaperdoesntknow/SymbioticLM-14B is a 17.8-billion-parameter symbolic–transformer hybrid model developed by Convergent Intelligence LLC: Research Division, built upon the Qwen-14B base. It integrates high-capacity neural representation with structured symbolic cognition, featuring persistent memory, entropic recall, and self-organizing knowledge structures. This experimental research checkpoint is designed for advanced reasoning agents, symbolic math/code generation, and multi-step conversational agents with true memory, supporting a 32768-token context length.

Loading preview...

Overview

SymbioticLM-14B is a 17.8-billion-parameter symbolic–transformer hybrid model developed by Convergent Intelligence LLC: Research Division. It combines a Qwen-14B transformer backbone with structured symbolic cognition, featuring persistent memory and generative symbolic evolution. This experimental research checkpoint is designed for advanced reasoning tasks, research assistance, and symbolic math/code generation, offering a unique approach to cognitive reasoning.

Key Capabilities

  • Hybrid Architecture: Integrates a Qwen-14B transformer with specialized symbolic modules like ThoughtDynamicsLNN, LiquidThoughtProcessor, CrystallineProcessor, and HelicalDNAProcessor.
  • Persistent Memory: Features 4096 symbolic states retrieved using entropy and contextual similarity, enabling multi-step conversational agents with true memory.
  • Symbolic Cognition: Supports long-form symbolic theorem generation, proof planning, scientific dialogue, and reasoning in fuzzy or discontinuous problem domains.
  • Dream Mode: Includes a background symbolic simulation for open-ended cognition.
  • Discrepancy Calculus Foundation: Leverages Discrepancy Calculus for self-generating completeness and stability criteria in memory consolidation.

Good for

  • Developing advanced reasoning agents requiring true memory and symbolic processing.
  • Research in symbolic math and code generation, including theorem proving.
  • Creating scientific dialogue systems and symbolic simulations.
  • Exploring cognitive models that bridge neural and symbolic AI paradigms.

Finetunes
1 models