INDEX

Explanations

proper names related to the name "Andrews"

oai_token-act-pair · gpt-3.5-turbo

the name "Andrews" and related mentions in the context of news or events

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

ozyg

-0.84

naire

-0.81

icides

-0.80

atorial

-0.79

ately

-0.77

NING

-0.76

ounty

-0.75

crime

-0.75

orescent

-0.75

icidal

-0.75

POSITIVE LOGITS

shire

1.00

 Andrews

0.96

 Lans

0.86

sburg

0.85

boro

0.81

endor

0.80

issance

0.77

hurst

0.77

AFB

0.75

mort

0.70

Activations Density 0.005%

proper names related to the name "Andrews"

the name "Andrews" and related mentions in the context of news or events

No Comments

No Known Activations