Text Classification
PEFT
Safetensors
English
decision-model
system-one
calibration
typesafe
decision-circuits
lora
Instructions to use jbarney/circuit-1.7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use jbarney/circuit-1.7b with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
Commit History
-
v2.2: open-taxonomy choice; v2.1's counterfactual and injection training
Verified signaturebf3b1fd
-
Card: v2.0, multi and locate questions, served results
Verified signature7a847cb -
circuit-1.7b v2.0 weights: multi, locate and ask-next questions
Verified signature47c7813 -
v1.3 card
Verified signaturef1d93ca -
v1.3 weights: + hard tier (long policy, contracts, multi-hop) at 4,096 tokens
Verified signature7df6089
-
v1.2: options encoded side by side; answers do not depend on option order
Verified signaturef59a9bf
-
Card: report Brier, not KL, against a model that rounds its probabilities
Verified signature1c04145 -
v1.1: groundedness and tool-call families added to the training data
Verified signature4fdb3b4
-
card: Nimble on Cole's bench
Verified signatureeaaf7ce -
card: link the circuit repo
Verified signature458e306 -
card: Cole's bench row
Verified signaturee0a9919 -
circuit-1.7b v0.1: LoRA + pointer head on Qwen3-1.7B-Base
Verified signature9e570fb -
initial commit
Verified signature71b9eaf