INDEX

Explanations

references to technology products, specifically those related to a brand called "Gear"

oai_token-act-pair · gpt-3.5-turbo

mentions of the "Gear" product series

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

umbn

-0.78

venue

-0.72

CTR

-0.71

article

-0.67

 Mathemat

-0.65

udeb

-0.65

Klu

-0.64

emet

-0.60

 denomin

-0.60

 Nurs

-0.60

POSITIVE LOGITS

 Gear

1.16

Gear

0.98

 GEAR

0.93

 gear

0.84

bag

0.80

hart

0.77

gear

0.77

Feed

0.75

heimer

0.75

bits

0.75

Activations Density 0.010%

references to technology products, specifically those related to a brand called "Gear"

mentions of the "Gear" product series

No Comments

No Known Activations