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

dayName

-0.71

iggurat

-0.70

ship

-0.69

alg

-0.68

ozy

-0.66

erg

-0.66

enture

-0.66

éĹĺ

-0.65

_-

-0.65

iquid

-0.63

POSITIVE LOGITS

 elector

0.74

 deaf

0.65

 ridden

0.64

okin

0.63

 imperson

0.63

otin

0.62

ifled

0.61

 transc

0.60

lite

0.58

eno

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.