Project: https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/
This is a pruned variant of OpenAI's GPT-OSS-20B model, reduced to 11 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 math 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 | ~8.4B (pruned from 21B) |
| Original Experts per Layer | 32 |
| Pruned Experts per Layer | 11 |
| 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 | Math |
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 mathematics-focused model utilizes experts that exhibited strong performance on mathematical reasoning tasks from MMLU mathematics subjects and quantitative sections. These experts excel at mathematical computation, proof strategies, and logical reasoning.
The expert selection process utilized our comprehensive analysis of router activation patterns across multiple evaluation benchmarks:
By identifying experts that consistently activated for math tasks, we created this specialized model that maintains domain expertise while significantly reducing computational requirements from 32 to 11 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-8.4b-specialized-math-pruned-moe-only-11-experts",
torch_dtype=torch.bfloat16,
device_map="cpu",
trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-8.4b-specialized-math-pruned-moe-only-11-experts")
# Generate with the model
messages = [
{"role": "user", "content": "Solve this equation: 2x + 5 = 17. Show your work step by step."}
]
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-8.4b-specialized-math-pruned-moe-only-11-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-8.4b-specialized-math-pruned-moe-only-11-experts")
# Generate with the model
messages = [
{"role": "user", "content": "Solve this equation: 2x + 5 = 17. Show your work step by step."}
]
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 math tasks:
Question: Solve this equation: 2x + 5 = 17. Show your work step by step.
Response:
<|channel|>analysis<|message|>We need to solve 2x+5=17. Steps: subtract 5, divide by 2. Provide explanation.<|end|><|start|>assistant<|channel|>final<|message|>**Equation to solve**
\[
2x + 5 = 17
\]
---
### Step 1: Isolate the term containing the variable
Subtract the constant term from both sides of the equation.
\[
2x + 5 - 5 = 17 - 5
\]
\[
2x = 12
\]
---
### Step 2: Solve for \(x\)
Divide both sides by the coefficient of \(x\) (which is 2).
\[
\frac{2x}{2} = \frac{12}{2}
\]
\[
x = 6
\]
---
### Result
\[
\boxed{x = 6}
\]
**Check**:
\(2(6) + 5 = 12 + 5 = 17\), which matches the original equation.<|return|>
Question: What is the Pythagorean theorem and how is it used?
Response:
<|channel|>analysis<|message|>Need explain theorem, usage.<|end|><|start|>assistant<|channel|>final<|message|>**Pyth‑theorem**
- **Statement**: In a triangle with sides *a*, *b*, *c* (where *c* is the longest side, the “hyp‑side”), the square of the longest side equals the sum of the squares of the other two sides.
\[
a^{2}+b^{2}=c^{2}
\]
- **Why**: It follows from the geometry of a triangle (area, angles) or from the algebraic definition of a distance in a 2‑dimensional space (the “Euclidean distance” in a vector space).
---
## How it is used
| **Field** | **Application** | **Key idea** |
|---|---|---|
| **Geometry** | Finding the lengths of a triangle (given two sides, find the third). | Rearsolve \(c=\sqrt{a^{2}+b^{2}}\). |
| **Engineering** | Calculing the height of a bridge, the width of a wall, or the length of a cable. | Use the formula to find the required dimensions. |
| **Computer graphics** | Determing the distance between two points in a 2‑D space. | Distance \(d=\sqrt{(x_2-x_1)^2+(y_2-y_1)^2}\). |
| **Physics** | In a 2‑D or 3‑D system, the conservation of energy or momentum (e.g., \(E=mc^{2}\)). | The “energy” is proportional to the square of the distance. |
| **Mat‑analysis** | Pro proving that the triangle inequality holds, or proving that the “Euclidean norm” is a metric. | Use the Pyth‑theorem as a basis for the metric’s properties. |
---
### Quick example
Supp you have a triangle with sides 3 cm and 4 cm.
Find the longest side:
\[
c=\sqrt{3^{2}+4^{2}}=\sqrt{9+16}=\sqrt{25}=5\text{ cm}
\]
So the triangle is 3‑4‑5, a classic “Pyth‑theorem” triangle.
---
**Bottom line**: The Pyth‑theorem is the “law of the squares” for a triangle (or a 2‑D space). It is used whenever
Question: Calculate the area of a circle with radius 7 meters.
Response:
<|channel|>analysis<|message|>Need area = π r^2. r=7 => 49π. Provide numeric.<|end|><|start|>assistant<|channel|>final<|message|>The area of a circle is given by the formula
\[
A = \pi r^{2}
\]
where \(r\) is the radius.
For a radius of \(7\) meters:
\[
A = \pi (7\,\text{m})^{2}
= \pi \times 49\,\text{m}^{2}
= 49\pi \,\text{m}^{2}
\]
So the area is **\(49\pi\) square meters**, which is approximately:
\[
49\pi \approx 49 \times 3.14159 \approx 154.16 \text{ m}^{2}
\]<|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}
}