INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ulhu

-0.68

hers

-0.67

odder

-0.65

 Annotations

-0.62

 Colonel

-0.62

 McCabe

-0.61

 Shroud

-0.61

 Chambers

-0.61

 cess

-0.60

 Engineers

-0.59

POSITIVE LOGITS

wcsstore

0.76

opian

0.74

GN

0.70

yg

0.69

yon

0.69

WARE

0.67

Trade

0.66

TOR

0.65

íķ

0.64

Lear

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.