INDEX

Explanations

references to the audience or viewers of a program

oai_token-act-pair · gpt-3.5-turbo

references to audience viewership statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Bain

-0.66

mine

-0.66

ulic

-0.64

 contingency

-0.63

 Bicycle

-0.63

 Surgery

-0.61

 Scrib

-0.61

 Shotgun

-0.61

 Bust

-0.60

 Plum

-0.60

POSITIVE LOGITS

 tuned

0.97

 viewers

0.93

 viewership

0.86

 tuning

0.81

wat

0.76

 viewing

0.76

 Ratings

0.74

akens

0.74

 watching

0.74

 feeds

0.73

Activations Density 0.028%

references to the audience or viewers of a program

references to audience viewership statistics

No Comments

No Known Activations