INDEX

Explanations

phrases related to issuing commands or making statements that assert control or authority

oai_token-act-pair · gpt-3.5-turbo

possessive pronouns and references to personal ownership

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

ozy

-0.72

ãĥ´ãĤ¡

-0.69

phase

-0.69

yang

-0.68

ablishment

-0.66

amia

-0.65

bender

-0.64

sequence

-0.64

ukong

-0.64

nown

-0.63

POSITIVE LOGITS

own

1.60

 arms

1.02

 sights

1.01

 feet

1.00

 belongings

0.98

 entire

0.98

 fingers

0.97

 knees

0.97

 hands

0.95

 gaze

0.95

Activations Density 0.151%

phrases related to issuing commands or making statements that assert control or authority

possessive pronouns and references to personal ownership

No Comments

No Known Activations