INDEX

Explanations

phrases related to continuing to read an article

oai_token-act-pair · gpt-3.5-turbo

phrases related to the act of continuing or reading content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ĪĴ

-0.88

ounty

-0.71

 bounty

-0.64

axter

-0.62

 Provincial

-0.60

senal

-0.59

eele

-0.59

issan

-0.58

vironment

-0.58

lain

-0.57

POSITIVE LOGITS

 reading

0.89

 Continue

0.66

advertisement

0.65

 Reading

0.62

flush

0.60

 CONTIN

0.60

 reader

0.59

 Comics

0.59

push

0.58

 Rivera

0.58

Activations Density 0.012%

phrases related to continuing to read an article

phrases related to the act of continuing or reading content

No Comments

No Known Activations