INDEX

Explanations

military and combat-related terms

oai_token-act-pair · gpt-3.5-turbo

military-related terminology and descriptions of operations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Seattle

-0.91

nect

-0.82

 Nerd

-0.81

phia

-0.80

Hamilton

-0.79

 Cinderella

-0.77

mson

-0.76

uckland

-0.75

ãĤ´ãĥ³

-0.75

Consumer

-0.74

POSITIVE LOGITS

 paramilitary

1.46

 militia

1.46

 mercenaries

1.40

 snipers

1.39

 militias

1.38

 weapons

1.38

 militants

1.36

 guerrilla

1.34

 hostages

1.31

 civilians

1.30

Activations Density 0.620%

military and combat-related terms

military-related terminology and descriptions of operations

No Comments

No Known Activations