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

mble

-0.82

externalActionCode

-0.82

EStream

-0.80

ãĤ´ãĥ³

-0.78

aber

-0.73

²¾

-0.72

gered

-0.71

eman

-0.71

aukee

-0.70

cot

-0.70

POSITIVE LOGITS

autions

0.72

 Bars

0.69

 sections

0.68

CFR

0.67

 provisions

0.67

 pages

0.67

 bars

0.65

scrib

0.64

RS

0.64

 Buch

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.