INDEX

Explanations

references to interfaces in programming, particularly in a generic or abstract context

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

identifier prefixes or suffixes

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

 pinulongan

-0.79

Tikang

-0.77

 defaultstate

-0.71



-0.71

脚注の使い方

-0.66

 дописавши

-0.66

 '\\;'

-0.65

 يتيمه

-0.65

 تضيفلها

-0.62

EndContext

-0.62

POSITIVE LOGITS

MockBean

0.38

 IERC

0.37

InjectAttribute

0.36

Certain

0.32

Les

0.32

 certain

0.31

ddle

0.30

addObject

0.30

 Cathy

0.29

 عين

0.29

Activations Density 0.001%

references to interfaces in programming, particularly in a generic or abstract context

identifier prefixes or suffixes

No Comments

No Known Activations