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

 cybersecurity

-0.69

 undet

-0.68

 omission

-0.64

 Canaver

-0.63

 coron

-0.63

osit

-0.62

 antiv

-0.62

 goose

-0.61

oy

-0.60

 pill

-0.60

POSITIVE LOGITS

bles

0.83

yles

0.76

bs

0.72

aved

0.71

ced

0.71

bl

0.70

bling

0.69

avage

0.69

ensional

0.68

bled

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.