INDEX

Explanations

references to parts of a whole or entities in a collective context

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

each of the

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

Portale

-0.54

stdlib

-0.50

omaterials

-0.47

thylamine

-0.46

 Reminis

-0.45

createObject

-0.45

postmedia

-0.44

orari

-0.43

Bibliographie

-0.42

$+\

-0.42

POSITIVE LOGITS

 ciascuno

0.66

 each

0.64

 mỗi

0.59

 EACH

0.59

Each

0.58

 Each

0.58

 ciasc

0.56

IntoConstraints

0.56

 chacune

0.56

DoubleQuotes

0.55

Activations Density 0.035%

references to parts of a whole or entities in a collective context

each of the

No Comments

No Known Activations