INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

reb

-0.87

sov

-0.84

chi

-0.82

roc

-0.82

 confir

-0.79

udget

-0.77

 reopen

-0.70

theless

-0.69

rial

-0.68

intosh

-0.68

POSITIVE LOGITS

 Niet

0.80

 Gates

0.79

 Ruler

0.76

 Destiny

0.75

 Ware

0.74

 Goddess

0.72

 Guardians

0.71

 Bungie

0.69

 Directive

0.68

 Mistress

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.