INDEX

Explanations

specific named entities and terms related to physical locations or contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_10/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Copeland

-0.50

<h3>

-0.49

 customerId

-0.46

mè

-0.46

toHaveBeenCalled

-0.45

//----

-0.45

 Guan

-0.45

endregion

-0.44

ClCompile

-0.44

 guan

-0.44

POSITIVE LOGITS

 BoxFit

0.61

Kathryn

0.57

 Schulze

0.56

 Eldridge

0.55

backdrop

0.54

 Dutton

0.54

 Barros

0.53

 ricevere

0.53

[*]

0.53

 Canvas

0.53

Activations Density 3.121%

specific named entities and terms related to physical locations or contexts

No Comments

No Known Activations