INDEX

Explanations

phrases related to specific instructions or steps in a process

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions or steps in a process

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

cedented

-0.86

ãĤ¦ãĤ¹

-0.77

Said

-0.74

ãĤ´ãĥ³

-0.70

SpaceEngineers

-0.67

ELD

-0.67

Į

-0.67

DeliveryDate

-0.66

çīĪ

-0.66

ãģ®æ

-0.64

POSITIVE LOGITS

try

1.33

 choose

1.25

 make

1.22

 consider

1.22

 prepare

1.22

 prioritize

1.21

 ensure

1.20

 visualize

1.20

 remember

1.19

ask

1.18

Activations Density 0.287%

phrases related to specific instructions or steps in a process

phrases indicating actions or steps in a process

No Comments

No Known Activations