INDEX

Explanations

elements related to characters' abilities and transformations

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

After specific words (enough, widely, equivalent, etc.)

np_acts-logits-general · gemini-2.0-flash

adverb or adjective follows

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 ligiloj

-0.43

 препратки

-0.41

Trend

-0.39

laquo

-0.38

 związane

-0.37

 <=",

-0.37

 electrolysis

-0.36

newList

-0.35

Morfologia

-0.35

issor

-0.35

POSITIVE LOGITS

GEBURTSDATUM

0.60

ViewFeatures

0.59

principalColumn

0.50



0.49

addGap

0.44

VersionUID

0.43

BuilderFactory

0.42

MigrationBuilder

0.41

RegressionTest

0.41

 AssemblyProduct

0.40

Activations Density 0.577%

elements related to characters' abilities and transformations

After specific words (enough, widely, equivalent, etc.)

adverb or adjective follows

No Comments

No Known Activations