INDEX

Explanations

negations and contractions related to belief or opinion

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

contractions with apostrophes

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

Kapcsolódó

-0.41

+#+#

-0.39

 UnityEditor

-0.37

 sikkert

-0.34

γέν

-0.32

bow

-0.32

stående

-0.31

 Einfluß

-0.31

Jesus

-0.31

 BioLib

-0.30

POSITIVE LOGITS

 JpaRepository

0.69

RenderAtEndOf

0.56

OGND

0.54

AutoScaleMode

0.54

Aholisi

0.53

ValueStyle

0.52

HtmlAttribute

0.52

AndEndTag

0.52

 חיצוניים

0.51

 Administrativna

0.51

Activations Density 0.050%

negations and contractions related to belief or opinion

contractions with apostrophes

No Comments

No Known Activations