INDEX

Explanations

proper nouns related to a specific person (Kapuscinski)

oai_token-act-pair · gpt-3.5-turbo

references to the name "Kapuscinski."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

urat

-0.86

eny

-0.78

ndra

-0.77

ename

-0.77

ources

-0.75

ereo

-0.74

quartered

-0.71

ensen

-0.71

cia

-0.70

agogue

-0.70

POSITIVE LOGITS

izen

0.88

 Okin

0.86

 Kang

0.79

ILA

0.75

 Extras

0.74

ã

0.71

Ki

0.71

@@@@@@@@

0.70

stall

0.69

à

0.69

Activations Density 0.028%

proper nouns related to a specific person (Kapuscinski)

references to the name "Kapuscinski."

No Comments

No Known Activations