INDEX

Explanations

system-related terms and references in code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

ftagPool

-0.44

pushFollow

-0.41

 nakalista

-0.41

 Италијани

-0.39

 ISNI

-0.37

rungsseite

-0.37

RotationOrder

-0.37

ніципа

-0.35

Publik

-0.34

 stras

-0.34

POSITIVE LOGITS

disposing

0.61

 CURIAM

0.55

astify

0.47

 judiciaire

0.46

/*

0.44

 useAppContext

0.44

 lenker

0.43

 sanitarias

0.42

IsContent

0.42

emma

0.41

Activations Density 0.000%

system-related terms and references in code

No Comments

No Known Activations