INDEX

Explanations

decimal numbers

oai_token-act-pair · gpt-3.5-turbo

all variations of numerical values, particularly those associated with specifications or measurements

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

 iceberg

-0.62

 advis

-0.56

 affili

-0.55

 numbered

-0.55

 overnight

-0.53

 overfl

-0.53

 worsh

-0.52

 accus

-0.50

hens

-0.49

 rabbit

-0.49

POSITIVE LOGITS

1.10

1.00

0.95

0.92

0.89

0.88

0.87

jpg

0.87

0.86

Activations Density 0.041%

decimal numbers

all variations of numerical values, particularly those associated with specifications or measurements

No Comments

No Known Activations