INDEX

Explanations

phrases describing reactions or responses to events or announcements

oai_token-act-pair · gpt-3.5-turbo

references to reactions or responses to various announcements or events

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

agate

-0.80

 Investigators

-0.69

 theorem

-0.67

 entrusted

-0.67

 Tycoon

-0.63

 Grimoire

-0.63

estate

-0.61

 Freed

-0.60

hart

-0.59

 Transactions

-0.58

POSITIVE LOGITS

 applause

1.49

 cheers

1.30

 reaction

1.29

 response

1.26

 disbelief

1.26

 reply

1.26

 laughter

1.24

 scorn

1.24

 backlash

1.23

 rebuke

1.22

Activations Density 0.276%

phrases describing reactions or responses to events or announcements

references to reactions or responses to various announcements or events

No Comments

No Known Activations