INDEX

Explanations

personal names

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter "L"

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

 Subtle

-0.73

 prohibition

-0.72

 Prohibition

-0.70

 boundaries

-0.70

 compensated

-0.65

 spirited

-0.62

 Hole

-0.61

grass

-0.61

 buck

-0.60

 correctness

-0.60

POSITIVE LOGITS

incoln

1.41

yrics

1.40

ONDON

1.37

opez

1.30

aptop

1.26

ateral

1.26

ifestyle

1.26

ongevity

1.24

IFE

1.23

adies

1.21

Activations Density 0.038%

personal names

occurrences of the letter "L"

No Comments

No Known Activations