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

tail

-0.73

su

-0.73

tal

-0.72

ele

-0.69

enough

-0.69

christ

-0.69

credit

-0.67

tails

-0.66

ozy

-0.66

len

-0.66

POSITIVE LOGITS

izoph

0.77

edia

0.74

heric

0.70

Eps

0.68

anooga

0.64

 Interactive

0.63

ileaks

0.62

ysc

0.62

undown

0.61

vernment

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.