You can deploy this model and start earning money today!

atenareply/
gemma-4-12b-asterion-agentic

Name: gemma-4-12b-asterion-agentic
Availability: InStock

Views

No views yet

Gemma-4-12B Asterion — Agentic

Verifiable maintenance assistant at 12B: telemetry anomaly triage + native Gemma-4 tool-calling + maintenance loop + raw-log reading. LoRA SFT on the ParamΔ model, dataset v2 (confound-free).

Overview

Stage: Agentic SFT (LoRA, merged) — dataset v2
Lineage: gemma-4-12B → CPT → ParamΔ → Agentic SFT v2 (this model)
Method: LoRA SFT (r=32/α=64, all-linear; merged) [lora] on grounded-by-construction tool trajectories [apigen] rendered with the NATIVE Gemma-4 template (assistant mask computed by char-offset span matching — no hand-written training template, perfect train/inference conditioning fidelity). Dataset v2 fixes: raw-vs-clean tool output decoupled from difficulty (the v1 confound behind the 1.2B's L2 recall dip) and Track-C multi-QA (4 typed questions per log). ≥50% real/teacher-anchored data [collapse].
Domain: fictional — Asterion Space Operations (24-satellite constellation: alert triage, anomaly investigation, subsystem ops) + reused Mars Express telemetry.

Training


Dataset	asterion-agentic-sft (v2: raw/clean decoupled, Track-C multi-QA)
Adapter	LoRA r=32, α=64, dropout 0.05, target all-linear (merged at push)
LR / epochs	2e-4 cosine, warmup 0.05, 1 epoch, max_len 4096, eff_batch 32
Tools	8 (same schema as the LFM2 round; Gemma-4 native wire format)

Evaluation

Metric	Value	Note
alert pass@1 (v3 verifier)	0.132	1.2B sibling: 0.304 — but see the profile
multi-alert SUBCASES by difficulty	0.22 / 0.22 / 0.22 / 0.12	L3/L4 move off zero with NO multi-alert training data — capability emerging with scale (1.2B: 0.00)
triage recall / precision (genuine)	0.68 / 0.56	over-calls genuine
distractor resistance	0.32	falls for correlated decoys
schema / grounding / budget	1.00 / 1.00 / 1.00	wire format is perfect
extraction pass@1	0.90	0.96 per-question
SFT eval_loss	0.285	0.381 → 0.285 over 1 epoch, no overfit

Eval is a held-out, deterministic verifiable harness (synthetic tasks); baseline = the pre-SFT ParamΔ model. See noval-corp/scripts/eval_agentic.py.

Intended use & limitations

Satellite-operations assistant: classify an ALERT (routine vs genuine) and run the maintenance loop on genuine anomalies. Use INSIDE a tool-execution loop (the model emits <|tool_call> blocks and consumes <|tool_response> results).

Limitations:

Tool-calling REQUIRES a multi-turn loop that executes tools and feeds results back.
Fictional domain; the genuine class is largely synthetic → eval measures verifier-conformance, not real-world detection.

Innovations tested

Native-template offset masking — instead of re-implementing Gemma-4's ~200-line chat template for a {% generation %} mask, training renders with the NATIVE template and locates the model-emitted spans (thought / tool calls / final answer) by char offsets.
Dataset v2 confound fix — raw-vs-clean evidence format is an independent per-task coin flip, verified statistically in the selfcheck — the format can no longer leak the label.

Usage

Use inside a tool-execution loop (the model emits calls; you run the tools and feed results back):

python

from transformers import AutoModelForCausalLM, AutoTokenizer

tok = AutoTokenizer.from_pretrained("atenareply/gemma-4-12b-asterion-agentic")
model = AutoModelForCausalLM.from_pretrained("atenareply/gemma-4-12b-asterion-agentic")
TOOLS = [...]  # the 8 OpenAI-style tool schemas (see noval-corp/config/curation.py: TOOL_SCHEMA)
messages = [{"role": "system", "content": "You are an Asterion operations assistant."},
            {"role": "user", "content": "ALERT on NPWD2531: 1.12A. Routine or genuine?"}]
inputs = tok.apply_chat_template(messages, tools=TOOLS, add_generation_prompt=True, return_tensors="pt", return_dict=True)
# parse <|tool_call>call:name{...}<tool_call|>, execute, append the result as a {"role":"tool",...}
# message WITH structured tool_calls on the assistant message, repeat.

Reference loop: noval-corp/scripts/eval_agentic_gemma.py::run_episode (wire spec: docs/gemma_wire_format.md).

Citations

LoRA — Hu et al., LoRA: Low-Rank Adaptation of LLMs, 2021. arXiv:2106.09685
APIGen / xLAM (execution-grounded tool data) — Liu et al., 2024. arXiv:2406.18518
GENIE (grounded synthesis) — Mirza et al., ICLR 2024. arXiv:2401.14367
Model collapse (≥50% real data) — Shumailov et al., 2024. arXiv:2404.01413
BFCL (AST function-call scoring) — Patil et al., ICML 2025
τ-bench (state-based agent eval) — Yao et al., 2024. arXiv:2406.12045

Card generated by noval-corp/scripts/gen_model_cards.py (standardized across the noval-corp model family).