Project: https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/
This is a pruned variant of OpenAI's GPT-OSS-20B model, reduced to 20 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 | ~13.7B (pruned from 21B) |
| Original Experts per Layer | 32 |
| Pruned Experts per Layer | 20 |
| 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 20 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-13.7b-specialized-math-pruned-moe-only-20-experts",
torch_dtype=torch.bfloat16,
device_map="cpu",
trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-13.7b-specialized-math-pruned-moe-only-20-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-13.7b-specialized-math-pruned-moe-only-20-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-13.7b-specialized-math-pruned-moe-only-20-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. Subtract 5: 2x=12. Divide 2: x=6. Provide step-by-step.<|end|><|start|>assistant<|channel|>final<|message|>**Step 1 – Write the equation**
\[
2x + 5 = 17
\]
**Step 2 – Isolate the term with \(x\)**
Subtract 5 from both sides so that the constant term disappears:
\[
2x + 5 - 5 = 17 - 5
\]
\[
2x = 12
\]
**Step 3 – Solve for \(x\)**
Divide both sides by 2 (the coefficient of \(x\)):
\[
\frac{2x}{2} = \frac{12}{2}
\]
\[
x = 6
\]
---
**Answer:** \(x = 6\)
(You can check by plugging back: \(2(6)+5 = 12+5 = 17\), which matches the right‑hand side.)<|return|>
Question: What is the Pythagorean theorem and how is it used?
Response:
<|channel|>analysis<|message|>Need explain theorem, formula, usage.<|end|><|start|>assistant<|channel|>final<|message|>**The Pythagorean Theorem**
> In a right‑angled triangle, the square of the length of the hypotenuse (the side opposite the right angle) equals the sum of the squares of the lengths of the other two sides.
Mathematically:
\[
a^{2} + b^{2} = c^{2}
\]
where
* \(a\) and \(b\) are the legs (the two sides that form the right angle),
* \(c\) is the hypotenuse.
---
### How It Is Used
| Application | What you do | Example |
|-------------|-------------|---------|
| **Finding a missing side** | Rear‑solve the equation for the unknown side | In a triangle with legs 3 m and 4 m, \(c = \sqrt{3^{2}+4^{2}} = 5\) m. |
| **Checking if a triangle is right‑angled** | Verify if \(a^{2}+b^{2}=c^{2}\) | For sides 6, 8, 10: \(6^{2}+8^{2}=36+64=100=10^{2}\) → right triangle. |
| **Design & construction** | Ensure structural elements meet right‑angle requirements (e.g., building corners, ladders). | A carpenter uses the theorem to calculate the length of a diagonal brace. |
| **Navigation & surveying** | Compute distances when only horizontal and vertical components are known. | A hiker measures 200 m east and 150 m north; the straight‑away distance is \(\sqrt{200^{2}+150^{2}}\). |
| **Computer graphics & animation** | Determine pixel distances, collision detection, or rendering of right‑angled shapes. | A program calculates the distance between two points in a 2‑D grid. |
| **Physics & engineering** | Resolve forces, velocities, or displacements into perpendicular components. | A projectile’s horizontal and vertical speeds combine to give the resultant speed. |
| **Mathematics & education** | Introduce geometry, algebra, and proof techniques. | Students prove that the area of a square built on each side of a right triangle satisfies the theorem. |
---
### Quick “Rule of Thumb”
- **If you know two sides** → compute the third with \(c = \sqrt{a^{
Question: Calculate the area of a circle with radius 7 meters.
Response:
<|channel|>analysis<|message|>Need area πr^2 = π*49 ≈ 153.938.<|end|><|start|>assistant<|channel|>final<|message|>The area \(A\) of a circle is given by
\[
A = \pi r^{2}
\]
where \(r\) is the radius.
For \(r = 7\) m:
\[
A = \pi (7\ \text{m})^{2}
= \pi \times 49\ \text{m}^{2}
\approx 3.14159 \times 49\ \text{m}^{2}
\approx 153.938\ \text{m}^{2}
\]
So the circle has an area of about **\(154\ \text{m}^{2}\)** (rounded to the nearest square meter).<|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}
}