INDEX

Explanations

phrases related to meetings or interactions

oai_token-act-pair · gpt-3.5-turbo

mentions of greetings or social meetings/interactions

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

 hourly

-0.85

 affected

-0.77

 hour

-0.75

 hardened

-0.72

 breed

-0.70

 total

-0.70

hus

-0.70

 oral

-0.70

 mediation

-0.70

 older

-0.69

POSITIVE LOGITS

high

1.32

miss

1.30

shoot

1.30

distance

1.30

prison

1.29

trade

1.29

goal

1.28

repeat

1.27

pret

1.27

advertising

1.26

Activations Density 0.062%

phrases related to meetings or interactions

mentions of greetings or social meetings/interactions

No Comments

No Known Activations