INDEX

Explanations

programming constructs and commands related to dynamic behavior and memory management

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

code keywords and identifiers

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

WebElementEntity

-0.51

 informée

-0.44

WriteTagHelper

-0.43

 détail

-0.40

 betweenstory

-0.39

OGND

-0.39

stateProvider

-0.37

 Layanan

-0.37

LookAnd

-0.37

 AspNetCore

-0.36

POSITIVE LOGITS

 Hann

0.51

NOSIS

0.51

 cheer

0.49

testify

0.49

arns

0.49

gano

0.48

gnu

0.48

rache

0.48

 hive

0.48

phous

0.48

Activations Density 0.004%

programming constructs and commands related to dynamic behavior and memory management

code keywords and identifiers

No Comments

No Known Activations