INDEX

Explanations

proper nouns, specifically related to the name "Nielsen"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Nielsen" along with related terms

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

 Carbuncle

-0.82

xual

-0.68

fare

-0.68

 Izan

-0.68

ebook

-0.66

course

-0.62

guards

-0.62

draw

-0.62

 Raider

-0.61

 hypot

-0.61

POSITIVE LOGITS

elsen

1.45

 Nielsen

0.99

Numbers

0.87

elson

0.80

obook

0.78

umbers

0.77

ott

0.75

wald

0.75

 "$:/

0.74

quist

0.74

Activations Density 0.004%

proper nouns, specifically related to the name "Nielsen"

mentions of the name "Nielsen" along with related terms

No Comments

No Known Activations