INDEX

Explanations

dates or periods of time

oai_token-act-pair · gpt-3.5-turbo

references to a particular time or season

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

rors

-0.66

ocity

-0.61

hess

-0.60

oller

-0.60

raq

-0.59

vre

-0.59

cks

-0.59

lessness

-0.57

gans

-0.57

 Coins

-0.57

POSITIVE LOGITS

 summer

1.29

 autumn

1.28

 year

1.26

 week

1.26

 weekend

1.25

 month

1.25

 fall

1.21

 coming

1.18

 winter

1.17

 spring

1.16

Activations Density 0.049%

dates or periods of time

references to a particular time or season

No Comments

No Known Activations