INDEX

Explanations

phrases related to secret operations or activities

oai_token-act-pair · gpt-3.5-turbo

references to social commentary or critique

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

ebin

-0.91

Rated

-0.88

Recommend

-0.87

enthal

-0.87

Solution

-0.85

Secondly

-0.81

`.

-0.81

Materials

-0.80

Dialogue

-0.79

****************

-0.78

POSITIVE LOGITS

 flashy

1.40

 glamorous

1.20

 sleek

1.17

 glam

1.14

 cheesy

1.12

 goofy

1.11

 slick

1.10

 fancy

1.10

 gimm

1.09

 glossy

1.09

Activations Density 0.840%

phrases related to secret operations or activities

references to social commentary or critique

No Comments

No Known Activations