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

olo

-0.73

auga

-0.73

uilt

-0.72

otos

-0.72

 veter

-0.71

 carbohyd

-0.71

eli

-0.70

acebook

-0.70

nces

-0.69

Services

-0.68

POSITIVE LOGITS

iqueness

0.62

 Paladin

0.60

fulness

0.60

 Dalai

0.59

 Rowling

0.59

 question

0.58

pot

0.58

 Burke

0.57

 Alexa

0.57

 issue

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.