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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hap

-0.87

ibur

-0.77

©

-0.72

 plateau

-0.71

stant

-0.70

ait

-0.69

ipolar

-0.68

 clauses

-0.68

 barg

-0.67

 horizont

-0.67

POSITIVE LOGITS

FUL

0.75

 Somali

0.74

Reviewer

0.73

Wiz

0.70

 Cullen

0.70

 Pilgrim

0.68

 Giov

0.67

 Ferr

0.67

CTV

0.66

MET

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.