INDEX

Explanations

phrases indicating the possibility or potential for action

oai_token-act-pair · gpt-3.5-turbo

expressions of possibility and suggestions for action

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

ò

-0.68

ñ

-0.66

ELD

-0.65

ukemia

-0.63

PLIED

-0.62

 subur

-0.62

ONSORED

-0.61

ikuman

-0.61

-0.60

conom

-0.60

POSITIVE LOGITS

 ourselves

1.49

 collectively

1.00

 incarcer

0.99

 discuss

0.99

 celebrate

0.96

 asses

0.95

 revisit

0.93

 eradicate

0.91

 adjourn

0.88

 hear

0.86

Activations Density 0.207%

phrases indicating the possibility or potential for action

expressions of possibility and suggestions for action

No Comments

No Known Activations