INDEX

Explanations

references to historical events or wars, particularly World War II

oai_token-act-pair · gpt-3.5-turbo

references to World War II

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

BOOK

-0.92

Tile

-0.82

notes

-0.81

 snipp

-0.80

TOP

-0.78

heet

-0.75

Tar

-0.75

Jar

-0.73

iances

-0.72

à

-0.72

POSITIVE LOGITS

 Memorial

0.93

era

0.88

hower

0.84

 Anniversary

0.80

 Veteran

0.79

 Bomber

0.78

 veterans

0.74

 Veterans

0.74

 bombers

0.73

 Syndrome

0.72

Activations Density 0.022%

references to historical events or wars, particularly World War II

references to World War II

No Comments

No Known Activations