INDEX

Explanations

keywords related to attracting attention or interest

oai_token-act-pair · gpt-3.5-turbo

instances of the word "attracted" 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Mandatory

-0.77

 Plain

-0.63

 error

-0.63

 timeout

-0.63

 iTunes

-0.61

 Pants

-0.60

 absolute

-0.59

 heavens

-0.58

 Standard

-0.58

ZA

-0.57

POSITIVE LOGITS

 attracted

3.65

 attracts

2.27

 attracting

1.98

 attract

1.94

 aroused

1.92

 lured

1.75

 attraction

1.55

 fascinated

1.52

 intrigued

1.44

 provoked

1.36

Activations Density 0.016%

keywords related to attracting attention or interest

instances of the word "attracted" in various contexts

No Comments

No Known Activations