INDEX

Explanations

comparative phrases that indicate magnitude or degree

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

 يتيمه

-0.49

 feroit

-0.44

 برانيه

-0.42

 BorderRadius

-0.41

 gouver

-0.38

 obé

-0.38

ɵɵelement

-0.37

NSUserDefaults

-0.36

 rév

-0.36

 démission

-0.36

POSITIVE LOGITS

 highly

0.88

 Highly

0.76

highly

0.74

Highly

0.72

 intensely

0.62

 altamente

0.62

 meest

0.59

 extremely

0.59

иболее

0.58

bardziej

0.57

Activations Density 0.356%

comparative phrases that indicate magnitude or degree

No Comments

No Known Activations