INDEX

Explanations

terms related to warfare, such as 'warfare', 'deterrence', and 'weaponry'

oai_token-act-pair · gpt-3.5-turbo

references to warfare and military strategies

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

icles

-0.78

gow

-0.76

umen

-0.74

este

-0.73

abet

-0.73

otto

-0.70

angelo

-0.69

val

-0.69

ocent

-0.68

ergy

-0.68

POSITIVE LOGITS

fare

1.07

rior

0.89

 Warfare

0.88

riors

0.85

 warfare

0.80

 hysteria

0.75

 waged

0.74

 Royale

0.73

 simulation

0.72

 battlefield

0.71

Activations Density 0.021%

terms related to warfare, such as 'warfare', 'deterrence', and 'weaponry'

references to warfare and military strategies

No Comments

No Known Activations