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

bard

-0.79

ILE

-0.65

cled

-0.65

eur

-0.65

tern

-0.64

nec

-0.64

erie

-0.64

amer

-0.64

inis

-0.64

oise

-0.63

POSITIVE LOGITS

 sqor

0.78

TPPStreamerBot

0.68

 dose

0.62

 McInt

0.61

Suk

0.59

 dividend

0.59

 dosage

0.59

 Haram

0.55

qv

0.55

Kul

0.55

Activations Density 0.000%

No Known Activations

This feature has no known activations.