INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

cale

-0.73

llah

-0.73

cius

-0.68

 attribution

-0.64

 unbiased

-0.63

 headphones

-0.61

ertodd

-0.60

isphere

-0.60

ilers

-0.58

OHN

-0.58

POSITIVE LOGITS

ãĥĥãĥī

0.79

wire

0.77

ãĥ´ãĤ¡

0.71

ãĤ¼

0.71

idated

0.69

ãĤ¨

0.68

SIGN

0.66

ãĥĳ

0.65

idine

0.65

ãĤ½

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.