INDEX

Explanations

references to military actions and conflicts

oai_token-act-pair · gpt-3.5-turbo

information regarding military actions and statements related to conflicts, particularly involving Russia and other countries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 cures

-0.97

enfranch

-0.88

 homelessness

-0.84

natureconservancy

-0.83

galitarian

-0.83

 addicts

-0.82

 addiction

-0.79

 alcoholism

-0.79

 Poverty

-0.79

 disenfranch

-0.79

POSITIVE LOGITS

 reconnaissance

1.39

 intercepted

1.24

 unmanned

1.24

 missile

1.23

USS

1.22

 airstrike

1.21

 aircraft

1.20

 Missile

1.16

 drone

1.16

 missiles

1.15

Activations Density 0.462%

references to military actions and conflicts

information regarding military actions and statements related to conflicts, particularly involving Russia and other countries

No Comments

No Known Activations