INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

anie

-0.73

ATIONAL

-0.69

wich

-0.67

ITH

-0.67

employment

-0.65

ulus

-0.63

intelligence

-0.63

ionage

-0.63

CIA

-0.62

enance

-0.61

POSITIVE LOGITS

rongh

0.69

ratom

0.69

 arteries

0.69

straight

0.67

 kilometres

0.66

iolet

0.65

 Abyss

0.64

 Piercing

0.63

ko

0.63

 sands

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.