INDEX

Explanations

mentions of advertisers, advertising, and compensation related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to advertising and advertisers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

chery

-0.77

ologies

-0.70

 COUR

-0.70

yon

-0.66

 Teach

-0.66

 Guam

-0.66

atche

-0.64

ttp

-0.62

Ã±a

-0.61

Ct

-0.61

POSITIVE LOGITS

 advertisers

1.14

 advertis

1.00

 marketers

0.87

 endors

0.85

vertis

0.82

 behavi

0.78

 advertising

0.77

vertising

0.77

 boycot

0.75

Premium

0.75

Activations Density 0.015%

mentions of advertisers, advertising, and compensation related terms

terms related to advertising and advertisers

No Comments

No Known Activations