INDEX

Explanations

elements related to HTTP requests and responses

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

code and international words

np_acts-logits-general · gemini-2.5-flash-lite

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

IBOutlet

-0.57

лтемелер

-0.56

íncia

-0.53

 Administrativna

-0.50

 Meksiku

-0.50

 <>",

-0.49

StructEnd

-0.49

 IBOutlet

-0.49

tidaknya

-0.46

 onCreateView

-0.45

POSITIVE LOGITS

Oru

0.37

 BoxDecoration

0.36

parsedMessage

0.35

0.34

Featured

0.34

tasche

0.33

 simplifié

0.32

Czytaj

0.32

 sappiamo

0.32

 informée

0.31

Activations Density 0.002%

elements related to HTTP requests and responses

code and international words

No Comments

No Known Activations