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

 Gleaming

-0.74

ovie

-0.73

lot

-0.72

 Greenberg

-0.72

SSL

-0.68

oleon

-0.67

SCP

-0.67

BuyableInstoreAndOnline

-0.66

 Goldberg

-0.66

Sed

-0.65

POSITIVE LOGITS

escription

0.93

ki

0.79

 padd

0.72

 referen

0.69

ado

0.69

 dere

0.68

 derby

0.68

 caut

0.67

 denomin

0.67

 rele

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.