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

hue

-0.68

 allegiance

-0.63

 recourse

-0.62

 Sadd

-0.61

 paradise

-0.60

ythm

-0.58

 Attribution

-0.58

 Unicorn

-0.56

FTWARE

-0.56

 fairy

-0.55

POSITIVE LOGITS

chell

0.79

thia

0.75

entin

0.71

heastern

0.70

ttes

0.69

ãĥĭ

0.68

zbek

0.68

reprene

0.67

colo

0.67

eki

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.