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

 casc

-0.68

Jac

-0.63

 unchecked

-0.62

 Pres

-0.61

 lair

-0.59

esc

-0.59

es

-0.58

 cascade

-0.58

 Chap

-0.58

 Klux

-0.58

POSITIVE LOGITS

UGC

0.91

WARE

0.80

0.77

furt

0.73

soDeliveryDate

0.72

SAY

0.70

umers

0.69

STEM

0.69

die

0.69

ISO

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.