INDEX

Explanations

quotations within text

oai_token-act-pair · gpt-3.5-turbo

quoted speech or dialogue

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 princip

-0.80

 deterrent

-0.78

£ı

-0.75

 proceeds

-0.73

 cannabin

-0.72

 flared

-0.70

 sund

-0.70

 conve

-0.70

 pole

-0.69

 viral

-0.69

POSITIVE LOGITS

Hey

1.32

Oh

1.25

hey

1.19

hello

1.18

Why

1.16

You

1.14

Sir

1.13

1.12

Eh

1.12

Yo

1.10

Activations Density 0.066%

quotations within text

quoted speech or dialogue

No Comments

No Known Activations