INDEX

Explanations

the word "subscriber" or its variations

oai_token-act-pair · gpt-3.5-turbo

references to subscriber counts and related metrics in digital services

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

 Pist

-0.69

ball

-0.67

Alz

-0.66

 Obst

-0.66

 Horror

-0.63

 Messi

-0.62

 Friedman

-0.62

Lob

-0.61

 Zurich

-0.61

wed

-0.61

POSITIVE LOGITS

 subscribers

1.12

hip

1.10

 subscriber

1.06

 subscriptions

0.94

 Unlimited

0.92

lio

0.87

 subscription

0.86

CHAT

0.80

 confir

0.78

riber

0.78

Activations Density 0.013%

the word "subscriber" or its variations

references to subscriber counts and related metrics in digital services

No Comments

No Known Activations