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

anon

-0.71

rises

-0.67

ictionary

-0.67

pace

-0.66

chwitz

-0.63

uzzle

-0.63

 vertex

-0.62

etitive

-0.62

enburg

-0.61

cript

-0.61

POSITIVE LOGITS

wav

0.70

ãģ®å®

0.69

DragonMagazine

0.67

eric

0.67

glers

0.63

so

0.62

shows

0.61

Wag

0.61

Dai

0.61

ļéĨĴ

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.