INDEX

Explanations

words related to drawing attention or reactions

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate significant attention, interest, or differences regarding social issues and events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

numbered

-0.66

 betrayed

-0.65

Õ

-0.65

 Inher

-0.64

 Atkins

-0.64

âķĲâķĲ

-0.62

cknow

-0.61

speak

-0.61

overed

-0.61

 Anat

-0.60

POSITIVE LOGITS

 attention

1.29

ire

1.19

riott

0.89

DragonMagazine

0.84

 wrath

0.82

MpServer

0.80

 Attention

0.80

 Flavoring

0.79

 conclusions

0.77

 crowds

0.76

Activations Density 0.176%

words related to drawing attention or reactions

phrases that indicate significant attention, interest, or differences regarding social issues and events

No Comments

No Known Activations