INDEX

Explanations

references to specific names or titles, particularly those containing 'Rod'

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

words ending in 'ers' or 'ily'

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 للاسماء

-0.59

 Económica

-0.56

 esternos

-0.52

 antemano

-0.50

%)$

-0.49

Spoljašnje

-0.48

 extranjera

-0.48

 colectiva

-0.47

ThroughAttribute

-0.47

 intere

-0.46

POSITIVE LOGITS

sol

0.52

der

0.44

synd

0.42

Yuc

0.42

lod

0.40

Fed

0.40

Lod

0.40

sch

0.39

nod

0.38

mud

0.38

Activations Density 0.025%

references to specific names or titles, particularly those containing 'Rod'

words ending in 'ers' or 'ily'

No Comments

No Known Activations