INDEX

Explanations

words related to receiving alerts, such as "alerts" and "notifications"

oai_token-act-pair · gpt-3.5-turbo

terms related to alerts and notifications

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

dx

-0.66

 Jinn

-0.63

knife

-0.63

atan

-0.62

EVA

-0.62

puff

-0.62

 relations

-0.62

ardless

-0.60

serving

-0.60

amins

-0.60

POSITIVE LOGITS

 Alert

1.01

 alerts

0.95

 alert

0.93

Alert

0.76

 notifications

0.74

mares

0.73

gency

0.72

 derail

0.69

ustration

0.69

ipolar

0.68

Activations Density 0.013%

words related to receiving alerts, such as "alerts" and "notifications"

terms related to alerts and notifications

No Comments

No Known Activations