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

©¶æ

-0.83

 Shutterstock

-0.73

 Canaver

-0.65

eland

-0.65

eless

-0.65

 Whitman

-0.64

rimination

-0.62

 fingert

-0.62

 Santana

-0.61

abama

-0.61

POSITIVE LOGITS

thouse

0.71

igel

0.64

Ð»

0.64

RL

0.63

 Editorial

0.62

ifest

0.61

CLA

0.61

Ly

0.60

etch

0.59

hai

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.