INDEX

Explanations

names of locations

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, specifically periods

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

orical

-0.77

imates

-0.73

INAL

-0.66

oric

-0.66

olate

-0.63

imus

-0.63

okin

-0.61

omed

-0.61

ictional

-0.60

 lifes

-0.60

POSITIVE LOGITS

CLASSIFIED

0.83

etc

0.79

etc

0.78

 EntityItem

0.74

soever

0.71

 Bethlehem

0.69

ternity

0.69

ooters

0.68

taboola

0.66

 Boone

0.66

Activations Density 0.030%

names of locations

punctuation marks, specifically periods

No Comments

No Known Activations