INDEX

Explanations

proper nouns related to people or characters

oai_token-act-pair · gpt-3.5-turbo

the repeated mention of the name "Hen" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

urable

-0.70

 bounce

-0.65

ãĥīãĥ©

-0.65

Activity

-0.65

ials

-0.64

olor

-0.63

::::::::

-0.63

sp

-0.62

ally

-0.62

Android

-0.61

POSITIVE LOGITS

Hen

3.87

Hen

2.95

 Henri

1.49

hen

1.23

Hun

1.23

 Hend

1.22

 Hutch

1.20

Wen

1.15

Hoo

1.10

 Freder

1.10

Activations Density 0.024%

proper nouns related to people or characters

the repeated mention of the name "Hen" in various contexts

No Comments

No Known Activations