INDEX

Explanations

phrases related to military operations or attacks

oai_token-act-pair · gpt-3.5-turbo

references to military operations and campaigns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

AppData

-0.92

rities

-0.90

Reason

-0.86

 Libraries

-0.84

 Painter

-0.80

riter

-0.76

 Authors

-0.75

models

-0.74

issues

-0.73

ween

-0.73

POSITIVE LOGITS

 raids

1.54

 raid

1.42

 attack

1.37

 attacks

1.33

 invasion

1.33

 bombing

1.33

 assaults

1.29

 bombardment

1.29

 assault

1.29

 airstrike

1.28

Activations Density 0.395%

phrases related to military operations or attacks

references to military operations and campaigns

No Comments

No Known Activations