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

çīĪ

-0.91

Ö¼

-0.83

lda

-0.74

å¥³

-0.68

 GOODMAN

-0.64

lled

-0.62

daq

-0.61

theless

-0.61

ãĥĺ

-0.60

 understatement

-0.60

POSITIVE LOGITS

iott

0.71

vant

0.69

bryce

0.65

 Nightmares

0.65

 Topic

0.64

Buzz

0.64

 Opposition

0.63

 Courier

0.62

hesive

0.62

lishes

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.