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

akis

-0.76

aft

-0.76

otte

-0.74

ithing

-0.70

odan

-0.70

acks

-0.70

owsky

-0.69

erc

-0.68

arov

-0.68

aughs

-0.65

POSITIVE LOGITS

LX

0.71

ħĭ

0.70

æ©Ł

0.65

quartered

0.65

 flyer

0.61

Alias

0.60

 weighted

0.59

 playbook

0.59

 datas

0.59

mble

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.