Downloads · 30 days
13
10% of all-time downloads
AmanPriyanshu/gpt-oss-10.8b-specialized-law-pruned-moe-only-15-experts
gpt-oss-10.8b-specialized-law-pruned-moe-only-15-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
13
10% of all-time downloads
All-time downloads
124
Public
Parameters
10.8B
21.5 GB on disk
Likes
1
Public
Click a slice to open those files.
.safetensors21.5 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 15 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 | ~10.8B (pruned from 21B) |
| Original Experts per Layer | 32 |
| Pruned Experts per Layer | 15 |
| 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 15 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-10.8b-specialized-law-pruned-moe-only-15-experts",
torch_dtype=torch.bfloat16,
device_map="cpu",
trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-10.8b-specialized-law-pruned-moe-only-15-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-10.8b-specialized-law-pruned-moe-only-15-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-10.8b-specialized-law-pruned-moe-only-15-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 the two main branches of the legal system that deal with very different kinds of disputes and wrongs. While they share some procedural rules (e.g., courts, evidence, appeals), they differ in purpose, parties, burden of proof, remedies, and consequences.
| Feature | Civil Law | Criminal Law |
|---|---|---|
| **Purpose** | Resolve disputes between private parties (individuals, businesses, government entities). | Protect society and enforce public order by puning conduct that harms the state or its citizens. |
| **Parties** | Plaintiff (the person or entity who claims a right was violated) vs. Defendant (the person or entity accused of wrongdoing). | State (or government) as prosecutor vs. Defendant (the accused). |
| **Burain of Proof** | “Pre‑probulo” or “balance of probabilities” – the plaintiff must show that it is more likely than unlikely that the defendant is liable. | “Beyond a reasonable doubt” – the prosecution must prove the defendant’s guilt to the highest standard. |
| **Remedies** | Typically compensatory: damages (mononey), injunctions, specific performance, or declarative judgments. | Penative: punishment (fines, community service, imprisonment, probation, or death penalty in some jurisdictions). |
| **Consequences for Defendant** | Civil liability may require payment or other actions, but the defendant remains free of criminal sanctions. | Criminal conviction can lead to loss of liberty, loss of civil rights, stigma, and collateral effects (e.g., employment, housing). |
| **Procedual Rights** | Defendant has a right to a fair trial, but the process is usually less adversarial and may be resolved by mediation or arbitration. | Defendant has extensive constitutional protections (e.g., Miranda rights, right to counsel, right to a jury, right to a speedy trial). |
| **Typical Cases** | Contract disputes, tort claims (negligence, defamation), property disputes, family law (division of assets, custody), landlord‑tenant disputes. | Theft, assault, fraud, homicide, drug offenses, white‑paper fraud, tax evasion, public‑office corruption. |
| **Evidence Rules** | Generally more flexible; hears‑out evidence may be admissible; the burden is lower. | Strict evidorial rules; hears‑out evidence is usually excluded; the prosecution must meet a higher
Question: Explain the concept of due process in legal systems.
Response:
<|channel|>analysis<|message|>Need concise explanation.<|end|><|start|>assistant<|channel|>final<|message|>**Due Process (or Due‑Process)** is a constitutional or legal principle that guarantees fair treatment of individuals by the state before any deprivation of life, liberty, or property. It is a safeguard against arbitrary or unjust government action and is found in many democratic legal systems, most notably in the U.S. Constitution (the 5th and 6th Amendments) and in international human‑rights instruments such as the Universal Declaration of Human Rights (Article 14) and the International Convention on the Protection of Human Rights and Fundamental Freedoms (Article 14).
---
### Core Elements
| Element | What it means | Typical application |
|--------|----------------|---------------------|
| **Notice** | The person must be informed of the action or charge against them. | A summons, indictment, or notice of a civil proceeding. |
| **Opportunity to Respond** | The person must have a chance to present evidence, argue, and cross‑examinate witnesses. | A trial, hearing, or administrative review. |
| **Judicial or Neutral Decision Maker** | The decision must be made by an impartial judge, jury, or tribunal. | Courts, tribunals, or administrative panels. |
| **Right to Counsel** | The person may be represented by a lawyer or have access to legal assistance. | Criminal defense, civil litigation, administrative appeals. |
| **Right to Appeal** | The decision can be reviewed by a higher court or body. | Appeals, habe‑corus, appellate courts. |
| **Equal Treatment** | The same rules apply to all, regardless of status or identity. | Non‑discrimination in procedural rights. |
---
### Why It Matters
1. **Pre‑deprivation** – It protects against the state taking away a person’s rights or property without a fair process.
2. **Transparency & Accountability** – It forces the state to justify its actions and allows oversight.
3. **Rule of Law** – It embodies the principle that government power is limited by law and must be exercised fairly.
---
### Variations Across Systems
| System | Key Features |
|--------|--------------|
| **U.S. Constitution** | 5th Amendment (due‑process before deprivation of life, liberty, property) + 6th Amendment (right to counsel, speedy trial). |
| **European Union** | Article 14 of the EU Charter of Fundamental Rights; national member states incorporate due‑process into domestic law
Question: What are the basic rights protected by the First Amendment?
Response:
<|channel|>analysis<|message|>Need to answer: freedom of religion, speech, press, assembly, petition.<|end|><|start|>assistant<|channel|>final<|message|>The First Amendment to the U.S. Constitution protects several core civil liberties. In short, it guarantees:
| Right | What it protects |
|---|---|
| **Freedom of Religion** | The right to practice any religion (or none) without government interference, and the prohibition of a state‑established religion. |
| **Freedom of Speech** | The right to express opinions, ideas, and viewpoints—whether spoken, written, or otherwise—without prior government censorship. |
| **Freedom of the Press** | The right for newspapers, radio, television, and other media to publish and broadcast content without government control. |
| **Right to Assemble** | The right for individuals to gather peacefully in public or private spaces for protest, demonstration, or other civic activities. |
| **Right to Petition** | The right to approach the government (or its representatives) to request action or redress. |
These five pillars—religion, speech, press, assembly, and petition—form the foundation of American free‑expression rights.<|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}
}