INDEX

Explanations

the word "Look"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "look" 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

icipated

-0.68

EStreamFrame

-0.65

 prolong

-0.65

âĹ¼

-0.63

adish

-0.61

iHUD

-0.58

"},"

-0.58

accompan

-0.57

riers

-0.57

estinal

-0.55

POSITIVE LOGITS

behind

0.74

ahead

0.73

 carefully

0.73

ers

0.72

outs

0.71

ye

0.70

 closely

0.70

etics

0.69

atio

0.69

holes

0.69

Activations Density 0.033%

the word "Look"

instances of the word "look" in various contexts

No Comments

No Known Activations