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

psons

-0.71

eva

-0.70

aiman

-0.68

ahead

-0.67

ott

-0.65

 grabs

-0.65

iscons

-0.63

behind

-0.61

ifi

-0.60

illery

-0.59

POSITIVE LOGITS

 Ritual

0.67

 Agric

0.67

 Chocobo

0.67

chwitz

0.66

0000000000000000

0.66

cffff

0.65

intestinal

0.64

activate

0.62

 Meadow

0.62

 Journals

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.