INDEX

Explanations

words related to specificity or uniqueness

oai_token-act-pair · gpt-3.5-turbo

phrases related to a specific focus or target

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

omore

-0.62

eni

-0.62

 thankfully

-0.61

 bombshell

-0.60

 withd

-0.58

 sadly

-0.57

reme

-0.56

etts

-0.55

 pioneered

-0.54

BuyableInstoreAndOnline

-0.54

POSITIVE LOGITS

 each

1.31

 whichever

1.05

 oneself

1.04

 your

1.02

 society

0.93

 their

0.90

each

0.82

 them

0.82

 respondent

0.82

 respective

0.80

Activations Density 0.489%

words related to specificity or uniqueness

phrases related to a specific focus or target

No Comments

No Known Activations