INDEX

Explanations

programming-related keywords and syntax

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

programming constructs

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

Източници

-0.73

SBATCH

-0.62

Tikang

-0.62

RenderAtEndOf

-0.60

Билгалдахарш

-0.59

 utafitiHapana

-0.49

expandindo

-0.48

endphp

-0.48

 beginnetje

-0.47

 newOwner

-0.46

POSITIVE LOGITS

mány

0.42

funktionen

0.40

全

0.35

매

0.34

Methods

0.34

vâ

0.33

 függ

0.32

Hook

0.32

yecek

0.32

ziehung

0.31

Activations Density 0.097%

programming-related keywords and syntax

programming constructs

No Comments

No Known Activations