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

 Scouting

-0.77

player

-0.72

 Healer

-0.69

ocular

-0.68

 ILCS

-0.66

Prev

-0.66

 Recomm

-0.65

bsp

-0.64

GPU

-0.63

 Goblin

-0.63

POSITIVE LOGITS

eny

0.74

 Strait

0.71

ande

0.64

rose

0.64

 beet

0.63

 Cosponsors

0.62

 complicit

0.61

AFP

0.60

eur

0.60

 cruc

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.