INDEX

Explanations

terms related to mathematical operations or calculations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

Accessory

-0.70

âĸ¬âĸ¬

-0.68

 Zimmer

-0.67

 Cosby

-0.65

ãĤŃ

-0.65

CLUD

-0.64

ãĥīãĥ©

-0.63

 Commodore

-0.63

STD

-0.62

 onward

-0.62

POSITIVE LOGITS

itled

1.33

racted

1.33

raction

1.31

lest

1.19

weet

1.17

otal

1.15

ript

1.12

ract

1.10

ribut

1.08

rop

1.08

Activations Density 0.002%

terms related to mathematical operations or calculations

No Comments

No Known Activations