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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hip

-0.74

peak

-0.71

 Logged

-0.67

hift

-0.67

Boost

-0.67

eper

-0.66

Posts

-0.66

uckland

-0.64

hoe

-0.64

 Aden

-0.64

POSITIVE LOGITS

¥µ

0.73

unal

0.71

 dissatisf

0.71

arsh

0.71

 totality

0.70

women

0.68

uably

0.66

 predec

0.64

é¾

0.63

 scrimmage

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.