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

Not in Any Lists

No Comments

Negative Logits

atching

-0.69

 endeavour

-0.69

bery

-0.69

itsu

-0.66

 alike

-0.65

quartered

-0.65

tons

-0.63

morrow

-0.62

brown

-0.62

 abduction

-0.62

POSITIVE LOGITS

Past

0.74

EStream

0.74

JSON

0.66

Asset

0.65

Writer

0.64

VE

0.64

Leave

0.64

Writ

0.64

Executive

0.64

EQ

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.