INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

kaya

-0.81

Shell

-0.72

Parser

-0.68

 Orion

-0.67

ulas

-0.66

ken

-0.66

Command

-0.66

 Osaka

-0.65

 relativity

-0.64

Parent

-0.64

POSITIVE LOGITS

 Removed

0.65

ydia

0.61

yip

0.61

oenix

0.60

 mourning

0.59

bel

0.59

cies

0.59

mount

0.59

 Next

0.59

 Cree

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.