INDEX

Explanations

articles and quantifiers related to quantities and locations

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

a/an followed by descriptive noun

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

ніципалі

-0.71

RegressionTest

-0.68

WebElementEntity

-0.65

MessageOf

-0.59

iettivo

-0.56

ViewFeatures

-0.55

 kasarigan

-0.54

 desmotivaciones

-0.53

kijken

-0.52

 veiks

-0.52

POSITIVE LOGITS

 highly

0.52

 very

0.51

 located

0.48

 rigid

0.45

 tough

0.43

 BoxDecoration

0.43

non

0.42

 extremely

0.41

 situated

0.40

 strong

0.38

Activations Density 0.013%

articles and quantifiers related to quantities and locations

a/an followed by descriptive noun

No Comments

No Known Activations