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

asse

-0.80

ories

-0.79

andem

-0.75

lez

-0.74

racted

-0.72

acles

-0.70

ropolitan

-0.69

redit

-0.69

amera

-0.69

inx

-0.68

POSITIVE LOGITS

nery

0.77

ļéĨĴ

0.66

holes

0.65

frames

0.64

 Neville

0.62

 STATS

0.62

BALL

0.61

 Brist

0.60

Stock

0.59

Row

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.