INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

hops

-0.78

 Wonderland

-0.69

 Tradable

-0.63

 WATCHED

-0.62

oway

-0.61

à¼

-0.59

rolet

-0.58

glers

-0.58

 Wond

-0.58

 assuming

-0.57

POSITIVE LOGITS

nces

0.86

nce

0.81

cv

0.79

ceptor

0.73

ISC

0.72

anim

0.71

vi

0.69

avis

0.68

Effect

0.67

bon

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.