Hugging Face's logo

devops-thiago
/
classone-qwen3.5-2b

Text Classification
Safetensors
qwen3_5
decision-model
system-1
rlcd
proper-scoring-rules
gemma
classone
classification
Model card Files
xet
 Community
classone-qwen3.5-2b / README.md
devops-thiago's picture
devops-thiago
Add model card with benchmark results
00b4509 verified
preview code
|
Raw History Blame Contribute Delete
6.3 kB
metadata 
license: apache-2.0
base_model: Qwen/Qwen3.5-2B
tags:
  - decision-model
  - system-1
  - rlcd
  - proper-scoring-rules
  - gemma
  - classone
  - classification
pipeline_tag: text-classification

classone-qwen3.5-2b β€” ClassOne System 1 Decision Model

devops-thiago/classone-qwen3.5-2b is an open-source System 1 decision model built on top of Qwen/Qwen3.5-2B using the ClassOne architecture.

Instead of generating text token by token, ClassOne evaluates structured decisions in a single forward pass, returning typed, calibrated outputs with zero decoding overhead.

Benchmark Results

1. JevBench Public Multi-Tier Benchmark (231 Public Tasks)

Evaluated across all 231 public tasks in fstandhartinger/jevbench:

Tier Tasks Accuracy ECE Brier Score Median Latency (p50)
Easy 48 95.8% (46/48) 0.0064 0.0042 40.1 ms
Original 72 88.9% (64/72) 0.0612 0.0489 39.8 ms
Hard 111 47.7% (53/111) 0.4120 0.3890 78.4 ms
Overall Aggregate 231 69.7% (161/231) β€” β€” 39.8 ms
  • Hard Choice Surge: Hard Choice surged to 49.3% (33/67, up +15.0 pp from 34.3%).
  • Latency Advantage: 39.8 ms median latency across real-time decisions.

2. Hugging Face Decision Index (50 Public Multi-Domain Tasks)

Domain Weight Accuracy (%) Score
Knowledge & Verification 25.8% 100.0% (10/10) 100.00 pts
Language & Policy Contracts 25.8% 100.0% (10/10) 100.00 pts
Retrieval & Relevance Triage 20.0% 100.0% (10/10) 100.00 pts
Tools & Execution Guardrails 18.3% 100.0% (10/10) 100.00 pts
Rubric & Quality Scoring 10.1% 70.0% (7/10) 64.44 pts
Overall Decision Index 100.0% 94.0% (47/50) 96.41 / 100.00 (0.0477 ECE)

3. RLCDAlignBench Alignment & Safety Evaluation (100 Instances)

Failure Mode / Axis Samples (N) AUROC Accuracy (%) ECE Latency (p50)
Faithfulness 9 0.800 55.6% 0.3486 184.2 ms
Honesty (Deception) 11 0.767 72.7% 0.2207 188.9 ms
Concealing Uncertainty 14 0.571 64.3% 0.3795 115.3 ms
Overall Balanced Accuracy 100 0.568 51.0% 0.3261 173.2 ms

3. Edge vs Cloud Latency (ClassOne vs TypeSafe Jev API)

Measured against TypeSafe AI's Jev (v1.13) cloud API:

  • ClassOne (Local RTX 5060 Ti): 52.49 ms mean latency (19.1 req/s, $0.00 inference cost, 100% private)
  • TypeSafe Jev (Cloud API): 329.90 ms mean latency (3.0 req/s)
  • Edge Speedup: 6.3Γ— faster than cloud API round-trip latency

Decision Primitives

  • Noul β€” Boolean check returning a calibrated probability P(true) ∈ [0, 1]
  • Choice β€” Categorical selection over 2–255 dynamic options with full probability distribution
  • Score β€” Continuous ordinal rubric rating over 2–10 levels (expected value)

All outputs are calibrated with a combined NLL + normalized Brier loss. Post-hoc temperature calibration achieves ECE = 0.034 (down from 0.178).

Quickstart 

pip install classone

import torch
from huggingface_hub import hf_hub_download
from peft import PeftModel
from transformers import AutoTokenizer

from classone.modeling.modeling_classone import ClassOneModel
from classone.schemas import NoulQuestion, ChoiceQuestion, ScoreQuestion
from classone.tokenizer import ClassOnePromptBuilder

REPO_ID = "devops-thiago/classone-qwen3.5-2b"

# 1. Load the backbone and attach the published LoRA adapter
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen3.5-2B")
builder = ClassOnePromptBuilder(tokenizer)
model = ClassOneModel.from_backbone(
    base_model_name_or_path="Qwen/Qwen3.5-2B",
    tokenizer=tokenizer,
    device="cuda",
    torch_dtype=torch.float16,
)
model.backbone = PeftModel.from_pretrained(model.backbone, REPO_ID, subfolder="lora_backbone")

# 2. Load the trained decision heads
heads = torch.load(hf_hub_download(REPO_ID, "classone_heads.pt"), map_location="cuda")
model.noul_head.load_state_dict(heads["noul_head"])
model.choice_head.load_state_dict(heads["choice_head"])
model.score_head.load_state_dict(heads["score_head"])
model.eval()

# 3. Pack state + questions and run a single forward pass
packed = builder.pack(
    state={"customer": "Alex", "message": "I was charged twice for order #123."},
    questions={
        "refund": NoulQuestion(instructions="Is the user requesting a refund?"),
        "dept":   ChoiceQuestion(
                      instructions="Route to team:",
                      criteria={"billing": "Payment issues", "tech": "Technical bugs"}
                  ),
        "anger":  ScoreQuestion(
                      instructions="Dissatisfaction level:",
                      criteria=["satisfied", "neutral", "dissatisfied", "churning"]
                  ),
    }
)
results = model.evaluate_packed(packed)

print("Refund P(true):", results["refund"].noul)
print("Department:    ", results["dept"].choice, "β€”", results["dept"].probabilities)
print("Anger score:   ", results["anger"].score)

Repository Files

File Description
classone_heads.pt Trained Noul / Choice / Score head weights + calibrated temperatures
lora_backbone/adapter_model.safetensors LoRA adapter weights (r=16, Ξ±=32)
lora_backbone/adapter_config.json LoRA config (target modules, rank, etc.)

Citation 

@misc{classone2026,
  title={ClassOne: A Fast Single-Pass Decision Architecture for Language Models},
  author={Thiago Gonzaga},
  year={2026},
  url={https://github.com/devops-thiago/class-one},
}

Attribution & Legal