INDEX

Explanations

terms related to mathematical calculations or estimations

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

oho

-0.80

heimer

-0.79

ĪĴ

-0.74

bley

-0.72

adder

-0.71

unin

-0.70

ici

-0.70

nered

-0.67

nen

-0.67

Bio

-0.67

POSITIVE LOGITS

 equations

1.09

 calculation

1.09

 calculations

1.06

 calculating

1.00

 equation

0.91

 probabilities

0.91

 Calcul

0.88

 calculated

0.87

 formulas

0.87

 estimating

0.86

Activations Density 0.070%

terms related to mathematical calculations or estimations

No Comments

No Known Activations