INDEX

Explanations

words related to a specific author

oai_token-act-pair · gpt-3.5-turbo

elements related to specific names and identifiers in a given context

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

 horr

-0.92

Fib

-0.85

 Alph

-0.82

Aff

-0.80

FUN

-0.77

Gon

-0.77

 reluct

-0.74

 Haram

-0.74

 Golden

-0.74

Alf

-0.74

POSITIVE LOGITS

ka

1.06

ke

1.05

cki

1.05

anski

1.03

aker

1.00

iken

0.93

akers

0.92

kee

0.92

zynski

0.91

owski

0.91

Activations Density 0.272%

words related to a specific author

elements related to specific names and identifiers in a given context

No Comments

No Known Activations