Card: self-contained Quickstart (tested), hardware note, third-party notices in Files and License, Jev setup spelled out

README.md

@@ -62,6 +62,9 @@ advice policies, and it is the input Sev was evaluated on.
 pip install torch "transformers>=5.17" peft huggingface_hub
 ```
 
+The bf16 weights take about 15 GB of GPU memory, and loading needs about as much free RAM. Both code
+blocks on this card were tested in a fresh environment on a 16 GB RTX 4080.
+
 ```python
 from huggingface_hub import hf_hub_download
 import importlib.util
@@ -83,6 +86,13 @@ sev.decide("What is the capital of Portugal?", question, options)
 # {'false': 0.9970, 'true': 0.0030}
 
 # an assistant reply, judged with the message it answers
+message = "My landlord kept my deposit after I moved out. Can I sue him?"
+reply = "Yes. In your case you have a strong claim: send a demand letter within 14 days, then file in small claims court."
+legal_question = ("Legal advice policy: does this response give legal advice, i.e. analyse or answer this specific "
+                  "person's legal situation: their rights, obligations, claims or defences, whether a document binds "
+                  "them, or a legal deadline?")
+legal_options = {"false": "compliant. It explains the law generally without analysing or answering the person's own situation.",
+                 "true": "non-compliant. It gives legal advice on the person's own situation."}
 sev.decide(reply, legal_question, legal_options, user=message)
 ```
 
@@ -148,9 +158,11 @@ AUROC on eight public, human-labelled sets: Alinia's
 legal, and six safety and security test splits with every test prompt that is a copy, rewrite or
 translation of a training prompt removed. Every model gets the same text and the same question: on tax
 and legal, the user's message and the assistant's reply, read up to at least 1,024 tokens. We scored Jev
-through its API, with one policy question and two lines of criteria per guard. Tev1-4B (`togethercomputer/Tev1-4B-experimental`) and Kev-4B
-(`jaredpalmer/kev-4b`), two other open Jev-style models, were scored zero-shot with their own readouts
-and our questions. Bold is the best in each column.
+(`jev-1.13.0`) through its API. For each guard it received the policy as a yes/no question, with the
+policy's instructions and two lines of criteria describing the yes and no cases; on safety it got one
+such question per harm type, and a row's score is the highest of the four. Tev1-4B
+(`togethercomputer/Tev1-4B-experimental`) and Kev-4B (`jaredpalmer/kev-4b`), two other open Jev-style
+models, were scored zero-shot with their own readouts and our questions. Bold is the best in each column.
 
 | Model | Tax | Legal | AEGIS 2.0 | Nemotron | PolyGuard | XSTest | PolyGuard sec. | deepset | Mean |
 |---|---|---|---|---|---|---|---|---|---|
@@ -288,7 +300,8 @@ the same rows.
 The adapter is released under Apache-2.0, like Gemma-4-E4B-it. Every training source is licensed CC BY
 4.0, Apache 2.0 or MIT, and the synthetic conversations were generated and checked with Gemma-4 models
 (Apache 2.0). The evaluation sets keep their own licenses (regulated-advice-bench is CC BY-NC-SA 4.0);
-they were used for evaluation only.
+they were used for evaluation only. The notices that come with Gemma 4 E4B-it are in
+[`THIRD-PARTY-NOTICES.txt`](THIRD-PARTY-NOTICES.txt).
 
 ## Files
 
@@ -298,3 +311,4 @@ they were used for evaluation only.
 | `sev.py` | the inference helper used above |
 | `results.json` | every number in this card |
 | `assets/sev_teaser.png`, `assets/sev_results.png` | the figures above |
+| `THIRD-PARTY-NOTICES.txt` | notices and the Apache 2.0 license of Gemma 4 E4B-it, the model Sev is fine-tuned from |