INDEX

Explanations

references to various national sports teams

oai_token-act-pair · gpt-3.5-turbo

references to national sports teams

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

DragonMagazine

-0.71

leys

-0.70

é»Ĵ

-0.69

ãĤ¨ãĥ«

-0.68

-0.66

Downloadha

-0.65

called

-0.64

DEP

-0.64

Meg

-0.64

Chart

-0.63

POSITIVE LOGITS

ICAN

0.81

 purse

0.81

 Museum

0.77

eers

0.70

 anthem

0.69

 Soccer

0.69

 honors

0.68

 unification

0.68

azeera

0.67

tv

0.67

Activations Density 0.049%

references to various national sports teams

references to national sports teams

No Known Activations