INDEX

Explanations

phrases related to exposing and documenting actions or behaviors

oai_token-act-pair · gpt-3.5-turbo

phrases related to academic documentation and discrimination issues

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

 GoPro

-0.89

operator

-0.78

 Kodi

-0.77

addons

-0.77

okemon

-0.75

yip

-0.75

 Cookies

-0.75

Operation

-0.75

 Garmin

-0.75

 Rabbit

-0.74

POSITIVE LOGITS

 professors

2.10

 academia

1.95

 faculty

1.91

 humanities

1.83

 universities

1.71

 academics

1.71

 undergrad

1.71

 graduate

1.70

 doctoral

1.70

 professor

1.68

Activations Density 0.706%

phrases related to exposing and documenting actions or behaviors

phrases related to academic documentation and discrimination issues

No Comments

No Known Activations