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

²¾

-0.73

leaf

-0.71

heter

-0.70

alysed

-0.69

 Ultr

-0.65

ket

-0.65

otom

-0.64

Plug

-0.64

thens

-0.63

lator

-0.63

POSITIVE LOGITS

 return

0.82

Ļ

0.78

orse

0.74

liest

0.71

acht

0.65

ega

0.65

emo

0.62

ILY

0.61

landish

0.60

ead

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.