Downloads · 30 days
18
13% of all-time downloads
AmanPriyanshu/gpt-oss-9.6b-specialized-law-pruned-moe-only-13-experts
gpt-oss-9.6b-specialized-law-pruned-moe-only-13-experts is a text generation model from AmanPriyanshu. Use it when you need the model to write or continue text. The card lists the license as apache-2.0.
Project: https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/
Downloads · 30 days
18
13% of all-time downloads
All-time downloads
137
Public
Parameters
9.6B
19.2 GB on disk
Likes
1
Public
Click a slice to open those files.
.safetensors19.1 GB · 100%
From the Hugging Face model README
Project: https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/
<div align="center">This is a pruned variant of OpenAI's GPT-OSS-20B model, reduced to 13 experts per layer based on activation patterns from the AmanPriyanshu/GPT-OSS-20B MoE Expert Activations dataset. We analyzed router decisions across evaluation benchmarks to identify and retain experts most relevant for law tasks.
⚠️ Experimental Model: This is an experimental pruned model that may not work well - check the examples below to see if the outputs meet your needs before use.
This pruning approach reduces the model size while attempting to preserve performance on the target domain.
| Metric | Value |
|---|---|
| Base Model | openai/gpt-oss-20b |
| Architecture | Mixture-of-Experts Transformer |
| Total Parameters | ~9.6B (pruned from 21B) |
| Original Experts per Layer | 32 |
| Pruned Experts per Layer | 13 |
| Layers | 24 |
| Top-k Routing | 4 |
| Context Length | 128K tokens |
| Attention Heads | 64 (Query), 8 (Key-Value) |
| Residual Dimension | 2880 |
| Attention Pattern | Alternating dense & sliding window (128 tokens) |
| Positional Encoding | RoPE (Rotary Position Embedding) |
| Normalization | RMSNorm |
| Precision | BF16 |
| License | Apache 2.0 |
| Specialization | Law |
Mixture-of-Experts models contain multiple specialized sub-networks (experts) per layer. During inference, only a subset of experts are activated for each token. Expert pruning involves:
Note: Performance may vary depending on how well the pruned experts match your specific use case.
This legal domain model employs experts that demonstrated expertise during law-related tasks from MMLU legal subjects. These experts excel at legal reasoning, jurisprudence, and understanding of legal frameworks and procedures.
The expert selection process utilized our comprehensive analysis of router activation patterns across multiple evaluation benchmarks:
By identifying experts that consistently activated for law tasks, we created this specialized model that maintains domain expertise while significantly reducing computational requirements from 32 to 13 experts per layer.
This model is based on analysis from the GPT-OSS-20B MoE Expert Activations dataset available at: 🔗 https://huggingface.co/datasets/AmanPriyanshu/GPT-OSS-20B-MoE-expert-activations
The dataset contains router activation patterns from OpenAI's GPT-OSS-20B model across diverse evaluation benchmarks, enabling the creation of these domain-optimized models through systematic expert pruning.
Our approach involves:
This is a direct pruning approach - no additional training was performed. The model inherits all capabilities from the original GPT-OSS-20B with focused expert selection.
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
# Load the specialized model on CPU
model = AutoModelForCausalLM.from_pretrained(
"AmanPriyanshu/gpt-oss-9.6b-specialized-law-pruned-moe-only-13-experts",
torch_dtype=torch.bfloat16,
device_map="cpu",
trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-9.6b-specialized-law-pruned-moe-only-13-experts")
# Generate with the model
messages = [
{"role": "user", "content": "What is the difference between civil and criminal law?"}
]
inputs = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt",
return_dict=True,
reasoning_effort="medium"
)
# Ensure inputs are on the same device as model
inputs = {k: v.to(model.device) for k, v in inputs.items()}
outputs = model.generate(
**inputs,
max_new_tokens=512,
do_sample=True,
temperature=0.1,
top_p=0.9,
pad_token_id=tokenizer.eos_token_id,
eos_token_id=tokenizer.eos_token_id
)
# Decode only the generated part
input_length = inputs['input_ids'].shape[1]
response_tokens = outputs[0][input_length:]
response = tokenizer.decode(response_tokens, skip_special_tokens=True)
print(response)
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
# Check MPS availability and load model
device = "mps" if torch.backends.mps.is_available() else "cpu"
model = AutoModelForCausalLM.from_pretrained(
"AmanPriyanshu/gpt-oss-9.6b-specialized-law-pruned-moe-only-13-experts",
torch_dtype=torch.float16, # Better MPS compatibility
device_map=device,
trust_remote_code=True,
low_cpu_mem_usage=True
)
tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-9.6b-specialized-law-pruned-moe-only-13-experts")
# Generate with the model
messages = [
{"role": "user", "content": "What is the difference between civil and criminal law?"}
]
inputs = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt",
return_dict=True,
reasoning_effort="medium"
)
# Move inputs to model device
inputs = {k: v.to(model.device) if isinstance(v, torch.Tensor) else v for k, v in inputs.items()}
# Use torch.no_grad for MPS stability
with torch.no_grad():
outputs = model.generate(
**inputs,
max_new_tokens=512,
do_sample=True,
temperature=0.1,
top_p=0.9,
pad_token_id=tokenizer.eos_token_id,
eos_token_id=tokenizer.eos_token_id,
use_cache=True
)
# Decode only the generated part
input_length = inputs['input_ids'].shape[1]
response_tokens = outputs[0][input_length:]
response = tokenizer.decode(response_tokens, skip_special_tokens=True)
print(response)
For CUDA GPU inference, the models work with minimal modifications from the CPU example above. Simply change:
device_map="auto" # Will automatically use GPU if available
torch_dtype=torch.bfloat16 # or torch.float16
The following examples demonstrate the model's performance on law tasks:
Question: What is the difference between civil and criminal law?
Response:
<|channel|>analysis<|message|>Need explain.<|end|><|start|>assistant<|channel|>final<|message|>**Civil law** and **criminal law** are two branches of the legal system that deal with very different kinds of disputes and wrongs. The main differences lie in the parties involved, the purpose of the law, the burden of proof, the remedies, and the procedural rules.
| Feature | Civil Law | Criminal Law |
|---|---|---|
| **Parties** | Private individuals or entities (e.g., a person suing a neighbor for a property dispute, a company suing another for breach of contract). | The state (or society) versus an individual or entity accused of violating a statute (e.g., theft, assault). |
| **Purpose** | To resolve disputes between parties and to compensate the wrong‑party (damers) for loss or injury. | To punish or deter wrongdoing that harms society, to protect public order, and to maintain social safety. |
| **Burain of Proof** | “Pre‑beyond a reasonable doubt” is not required; the plaintiff must prove the claim **by a pre‑balance of evidence** (i.e., “more likely than less”). | The prosecution must prove the defendant’s guilt **beyond a reasonable doubt**. |
| **Remade** | Typically monetary damages, injunctions, or specific performance (e.g., pay compensation, stop a nuisance). | Penative sanctions: fines, imprisonment, probation, community service, restitution, or other corrective measures. |
| **Procedure** | Usually civil courts or tribunals; parties may choose to represent themselves or hire counsel; the process is often less formal and can be faster. | Criminal courts (often with a judge and jury); the defendant has constitutional rights (e‑e. right to counsel, right to a speedy trial, right to a presis). The process is more formal and can be longer. |
| **Defender’s Rights** | The defendant (or “respondent”) has a right to a fair hearing, but the state does not have a special “prosecution” role. | The defendant has a right to a presis, to counsel, to cross‑examination, to a jury, and to appeal. |
| **Outcome** | The court orders the defendant to pay or to do something; the plaintiff receives compensation. | The court orders the defendant to pay a fine, serve jail time, or perform community service. |
| **Examples** | Contract disputes, property disputes, tort claims (neglig
Question: Explain the concept of due process in legal systems.
Response:
<|channel|>analysis<|message|>Need explain due process.<|end|><|start|>assistant<|channel|>final<|message|>**Due Process (or “Due‑Process”) – a Fundamental Legal Principle**
---
### 1. What it Is
* **Due Process** is a constitutional or statutory safeguard that guarantees a fair, impartial, and transparent legal procedure before the state may depriment an individual’s life, liberty, or property.
* It is a *procedural* right, not a substantive right. It protects the *process* by which a government acts, not the outcome itself.
---
### 2. Core Elements
| Element | Why It Matters | Typical Manifestation |
|--------|--------------|------------------------|
| **Notice** | The person must be informed of the claim or charge. | Written summons, court docket, public notice. |
| **Opportunity to Respond** | The person must be able to present evidence, argue, and cross‑examine. | Oral or written pleadings, deposition, trial. |
| **Neutral Decision Maker** | The decision must be made by an impartial judge or jury. | Ad‑un‑parte judge, jury trial. |
| **Right to Counsel** | The person may have legal representation. | Public‑law attorney, private counsel. |
| **Right to Appeal** | The decision can be reviewed by a higher court. | Appeal, habe‑de‑liba, appellate review. |
| **Reasoned Decision** | The outcome must be based on evidence and law. | Written opinion, record of findings. |
| **Timity** | The process must be timely, preventing indefinite detention or punishment. | Stat‑tim, bail, release. |
---
### 3. Historical Roots
| Era | Key Development |
|-----|----------------|
| **17th–18th Century** | English Common Law – “no law without the consent of the people.” |
| **19th Century** | U.S. Constitution (17th Amendment) – “no law shall be passed without the consent of the people.” |
| **20th Century** | 4th & 5th Amendments (U.S.) – “No person shall be deprived of life, liberty, or property … without due process.” |
| **International** | Universal Declaration of Human Rights (1945) – “Everyone has the right to a fair trial.” |
---
### 4. How It Works in Practice
1. **Arar‑diction** – The state must show a
Question: What are the basic rights protected by the First Amendment?
Response:
<|channel|>analysis<|message|>Need to answer: rights: freedom of religion, freedom of speech, press, assembly, petition, right to bear arms? Actually 1st Amendment: prohibits laws abrapping religion, speech, press, assembly, petition, and also right to bear arms. Provide explanation.<|end|><|start|>assistant<|channel|>final<|message|>The First Amendment to the U.S. Constitution protects a handful of fundamental civil liberties that are considered essential to a free, democratic society. In short, it guarantees:
| **Right** | **What it protects** |
|---|---|
| **Freedom of Religion** | The government may not establish a religion or prohibit the practice of any religion. |
| **Freedom of Speech** | Individuals may express opinions, ideas, and viewpoints without government censorship or punishment. |
| **Freedom of the Press** | The press (newspapers, radio, television, etc.) may publish information and critique the government. |
| **Freedom of Assembly** | People may gather peacefully for protest, demonstration, or other purposes. |
| **Freedom to Petition** | Citizens may request or demand action from the government. |
| **Right to Bear Arms** | (as interpreted by the Supreme Court) Individuals may possess and carry firearms. |
These rights are not absolute; the courts have carved out certain limits (e.g., restrictions on hate‑speech, defamation, obscusity, or threats of violence). But the First Amendment remains the cornerstone of American free‑speech, religious liberty, and democratic participation.<|return|>
If you use this model in your research, please cite:
@misc{priyanshu2025gptoss,
title={{GPT-OSS MoE Expert Fingerprinting: Analyzing Expert Activation Patterns in Mixture of Experts Models}},
author={Priyanshu, Aman and Vijay, Supriti},
year={2025},
howpublished={\url{https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/}},
note={Interactive analysis tool for expert activation patterns in MoE architectures}
}