INDEX

Explanations

phrases related to refusal or non-compliance with a request or action

oai_token-act-pair · gpt-3.5-turbo

phrases indicating refusal or resistance to performing actions

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

soDeliveryDate

-0.99

 Compass

-0.78

lined

-0.76

oen

-0.72

quickShipAvailable

-0.69

 Header

-0.69

Line

-0.68

////////////////////////////////

-0.68

retty

-0.67

GW

-0.67

POSITIVE LOGITS

 accept

1.18

 endorse

1.12

 participate

1.08

 cooperate

1.08

 reproduce

1.07

 comply

1.05

 obey

1.04

 recognize

1.04

 settle

1.04

 acknowledge

1.03

Activations Density 0.061%

phrases related to refusal or non-compliance with a request or action

phrases indicating refusal or resistance to performing actions

No Comments

No Known Activations