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

adh

-0.77

alan

-0.72

uberty

-0.72

iation

-0.71

uddin

-0.71

ylum

-0.70

aryn

-0.68

 suffice

-0.64

 Isis

-0.64

iations

-0.64

POSITIVE LOGITS

 opposite

0.69

doors

0.69

bags

0.66

lined

0.66

Wit

0.65

 again

0.63

ney

0.62

 Draper

0.62

eteria

0.62

 Boom

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.