INDEX

Explanations

terms related to data structures and byte manipulation

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

buffer / bytes

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

//

-0.48

<?

-0.42

Nationalité

-0.36

SpringBootTest

-0.35

ServletContext

-0.32

/**

-0.32

RESTRICT

-0.32

NamedQuery

-0.32

(({

-0.31

 robić

-0.31

POSITIVE LOGITS

bytes

0.89

 bytes

0.73

httphttps

0.71

Bibliograf

0.66

Tikang

0.66

ChildScrollView

0.65

ptest

0.65

ScopeManager

0.64

 protoimpl

0.64

 Bytes

0.62

Activations Density 0.001%

terms related to data structures and byte manipulation

buffer / bytes

No Comments

No Known Activations