INDEX

Explanations

phrases or terms that contain the word "we"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the pronoun "we."

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

 gratification

-0.65

 citation

-0.64

 firsthand

-0.63

 conflicts

-0.62

 contradictions

-0.61

 citations

-0.60

 Authority

-0.60

 Kush

-0.58

 contradiction

-0.57

 sectarian

-0.57

POSITIVE LOGITS

bsite

1.44

athered

1.28

eping

1.19

arers

1.16

aning

1.15

lder

1.14

aving

1.14

igh

1.13

eps

1.13

avers

1.12

Activations Density 0.085%

phrases or terms that contain the word "we"

occurrences of the pronoun "we."

No Comments

No Known Activations