INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

emetery

-0.77

 conflic

-0.74

senal

-0.68

 Gifts

-0.68

ems

-0.67

Naz

-0.67

 appro

-0.66

usalem

-0.65

 livest

-0.65

estial

-0.64

POSITIVE LOGITS

isen

0.75

pedia

0.73

strate

0.71

iston

0.69

iard

0.67

 Rolls

0.65

 lawy

0.65

porter

0.64

rock

0.64

izer

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.