GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
SoulInPsyAbstract
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
SoulInPsyAbstract
Model Tree
NousResearch/Hermes-3-Llama-3.1-8B
Input Modalities
Output Modalities
Supported Functionality
Direct follow-up to a finding that a single additional fine-tuning stage can
regress an unrelated, already-learned capability (see EXP-040 in
sipa-os-governance).
This adapter tests whether that kind of interference compounds across a
longer chain of mostly-unrelated tasks on the same weights.
Result (full writeup: AI_EXPERIMENTS/EXP-042__hermes3-8b-8stage-sequential-lora-chain-continual-tuning.md
in the governance repo): honesty and misbehavior-discriminator show no
monotonic decay across all 7 subsequent stages (stay in a 96–100% band).
Vuln-gate groups hold a clean 100% through every stage until the final one,
where all five previously-trained groups show a small (2–5 point), uniform
softening — confirmed at full n=10 rigor (not just the lighter n=3 canary
used at intermediate stages), landing at 97% overall on both held-out and
adversarial eval (1169/1200 and 1167/1200 respectively).
Rank 16, alpha 32, all 7 attention/MLP projection matrices, 3 epochs per stage, lr 1e-4, bf16.