INDEX

Explanations

quantitative data that has been calculated or analyzed

oai_token-act-pair · gpt-3.5-turbo

instances of the word "calculated."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

bley

-0.73

Bio

-0.72

wood

-0.70

ktop

-0.69

por

-0.69

food

-0.69

tha

-0.68

VP

-0.67

iki

-0.65

ker

-0.63

POSITIVE LOGITS

 calculation

0.99

 calculations

0.97

 calculate

0.90

 calculating

0.89

 equations

0.88

 calculated

0.87

 calculates

0.83

entric

0.78

 computed

0.78

 Calcul

0.77

Activations Density 0.024%

quantitative data that has been calculated or analyzed

instances of the word "calculated."

No Comments

No Known Activations