INDEX

Explanations

code-related elements and programming constructs

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

graphics programming terms

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

RegressionTest

-0.53

RefNanny

-0.51

期刊论文

-0.50

rophes

-0.48

#+#

-0.48

 unknownFields

-0.48

contentLoaded

-0.48

 CanadaChoose

-0.47

الإنجليزية

-0.47

EndContext

-0.47

POSITIVE LOGITS

AnimationsModule

0.44

PTSD

0.37

يكب

0.37

XHR

0.36

NFT

0.35

onInit

0.33

stateMutability

0.32

Aholisi

0.32

PPR

0.32

ereum

0.32

Activations Density 0.017%

code-related elements and programming constructs

graphics programming terms

No Comments

No Known Activations