INDEX

Explanations

dates written in a numerical format (MM/DD/YY) with high activation values

oai_token-act-pair · gpt-3.5-turbo

numerical data, particularly dates and quantities

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

essor

-0.69

ADVERTISEMENT

-0.63

agna

-0.60

Choice

-0.60

IES

-0.59

anta

-0.59

 grou

-0.59

una

-0.58

matically

-0.58

kas

-0.58

POSITIVE LOGITS

1.16

ths

1.03

/,

0.99

/"

0.94

/.

0.91

/_

0.90

/#

0.82

/)

0.79

/-

0.78

)=(

0.77

Activations Density 0.044%

dates written in a numerical format (MM/DD/YY) with high activation values

numerical data, particularly dates and quantities

No Comments

No Known Activations