INDEX

Explanations

terms related to locating or finding something

oai_token-act-pair · gpt-3.5-turbo

vocabulary related to searching and positioning

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

Ans

-0.73

weet

-0.72

 Decay

-0.67

 Comedy

-0.66

Raj

-0.64

 jokes

-0.62

Cry

-0.62

laughter

-0.62

 Mens

-0.61

 tant

-0.61

POSITIVE LOGITS

 locate

3.28

 locating

2.53

 relocate

1.94

 relocated

1.69

 located

1.61

 location

1.50

 relocation

1.40

loc

1.39

 Located

1.32

 Location

1.30

Activations Density 0.021%

terms related to locating or finding something

vocabulary related to searching and positioning

No Comments

No Known Activations