INDEX

Explanations

patterns related to code or file paths, particularly in the context of programming languages and libraries

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

file paths or code snippets

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

Réalisation

-0.55

jsxFileName

-0.55

 ModelExpression

-0.50

évaluateur

-0.49

 disambiguazione

-0.47

 jspb

-0.47

原始内容存档于

-0.47

IsContent

-0.47

queryInterface

-0.46

forChild

-0.44

POSITIVE LOGITS

Diagonal

0.43

itonitis

0.41

achery

0.40

tershire

0.40

yntaxException

0.40

vrt

0.39

 secours

0.39

honneur

0.38

nesto

0.38

zeption

0.38

Activations Density 0.013%

patterns related to code or file paths, particularly in the context of programming languages and libraries

file paths or code snippets

No Comments

No Known Activations