INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

orship

-0.82

ilee

-0.72

inition

-0.70

Å«

-0.66

initions

-0.65

Sem

-0.65

sembly

-0.65

EMENT

-0.64

 strand

-0.63

Commercial

-0.62

POSITIVE LOGITS

 Goblin

0.77

kas

0.76

é¾

0.66

pub

0.63

RM

0.63

ston

0.63

naire

0.63

 adren

0.62

tf

0.61

cdn

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.