INDEX

Explanations

personal pronouns and determiners followed by positive or negative opinions

oai_token-act-pair · gpt-3.5-turbo

the pronoun "it" and related expressions about it

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

interstitial

-0.66

 Grateful

-0.57

 commencement

-0.57

 Colleg

-0.56

Joined

-0.56

 Profession

-0.56

 Everyday

-0.55

£ı

-0.55

 Afterwards

-0.54

 Corrections

-0.53

POSITIVE LOGITS

 could

1.46

 wouldn

1.44

 would

1.40

could

1.35

 might

1.32

would

1.24

 certainly

1.24

 likely

1.22

 shouldn

1.20

 ought

1.20

Activations Density 0.530%

personal pronouns and determiners followed by positive or negative opinions

the pronoun "it" and related expressions about it

No Comments

No Known Activations