INDEX

Explanations

references to specific historical figures or concepts related to literature and political ideologies

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

historical and cultural descriptions

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

 ERSITY

-0.46

 pizarra

-0.44

 <<<<<<<<<<<<<<

-0.44

 Mulher

-0.43

NameInMap

-0.43

 favor

-0.40

badlogic

-0.40

 favour

-0.39

 favours

-0.39

 Catalana

-0.39

POSITIVE LOGITS

LookAnd

0.45

 Мексичка

0.42

 nargin

0.41

 setMessage

0.40

AnchorStyles

0.40

 CGRectMake

0.39

 հղումներ

0.39

༚

0.39

 getKey

0.39

 antis

0.39

Activations Density 0.030%

references to specific historical figures or concepts related to literature and political ideologies

historical and cultural descriptions

No Comments

No Known Activations