INDEX

Explanations

highly engaging or attention-grabbing video titles with strong emotional or sensational language

oai_token-act-pair · gpt-3.5-turbo

mentions of video content

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

arious

-0.81

bably

-0.68

ĺħ

-0.67

 Classification

-0.66

 Preservation

-0.64

omnia

-0.64

orescence

-0.63

Ļ½

-0.63

ongevity

-0.63

idia

-0.63

POSITIVE LOGITS

 VIDEOS

0.80

 IMAGES

0.76

iques

0.74

å§

0.71

natureconservancy

0.70

SEN

0.68

thumbnails

0.67

 than

0.64

utm

0.62

advertisement

0.61

Activations Density 0.023%

highly engaging or attention-grabbing video titles with strong emotional or sensational language

mentions of video content

No Comments

No Known Activations