INDEX

Explanations

references to the soccer club "Arsenal" and related terms

oai_token-act-pair · gpt-3.5-turbo

references to the football club Arsenal

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

obe

-0.77

aho

-0.74

umbn

-0.70

ommel

-0.69

pty

-0.68

margin

-0.66

states

-0.63

leading

-0.62

orders

-0.62

ebin

-0.62

POSITIVE LOGITS

Arsenal

0.88

 Skydragon

0.79

 Wenger

0.78

 Arsenal

0.78

FC

0.76

AFC

0.71

bane

0.70

 Yard

0.69

 mosqu

0.69

 Sports

0.69

Activations Density 0.009%

references to the soccer club "Arsenal" and related terms

references to the football club Arsenal

No Comments

No Known Activations