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

robe

-0.82

ynchron

-0.77

detail

-0.74

endment

-0.70

rop

-0.69

eport

-0.69

lock

-0.68

lus

-0.67

scan

-0.67

plates

-0.66

POSITIVE LOGITS

BuyableInstoreAndOnline

0.83

Bai

0.69

imum

0.69

 Beng

0.68

idia

0.66

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.64

argon

0.64

ERY

0.62

 antid

0.62

IVES

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.