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

spect

-1.02

 Blizz

-0.80

Tid

-0.77

-0.75

VIS

-0.72

adow

-0.71

Rated

-0.71

 Toys

-0.69

ey

-0.69

 Surprise

-0.68

POSITIVE LOGITS

netflix

0.72

 persecut

0.71

 operator

0.66

 prelim

0.66

 manif

0.66

hyde

0.65

 arbitrary

0.64

 soph

0.64

 arming

0.63

achus

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.