INDEX

Explanations

phrases related to enhancing user abilities or powers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

脚注の使い方

-0.71

 contextLoads

-0.68

=$?

-0.65

InitVars

-0.63

 ***!

-0.63

Билгалдахарш

-0.61

новниш

-0.60

AndEndTag

-0.57

windowFixed

-0.57

 μαύ

-0.56

POSITIVE LOGITS

 boost

0.55

 boosts

0.51

 increases

0.51

 campsite

0.51

0.48

 improves

0.47

 increase

0.47

 percentage

0.47

 propOrder

0.47

 enhanced

0.46

Activations Density 0.020%

phrases related to enhancing user abilities or powers

No Comments

No Known Activations