INDEX

Explanations

references to academic contributions or authors' work in scientific texts

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

work contribution

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

IVEREF

-0.57

providedIn

-0.54

Хьажоргаш

-0.52

antMatchers

-0.52

hpur

-0.51

AndroidJUnit

-0.51

obs

-0.50

XmlEnum

-0.50

bootstrapcdn

-0.50

Derbyniad

-0.49

POSITIVE LOGITS

 irmã

0.41

 žem

0.41

 gantung

0.40

ⓘ

0.40

 Decke

0.39

 beleza

0.39

 esfuerzan

0.39

 acompañado

0.38

 sekaligus

0.38

Also

0.38

Activations Density 0.000%

references to academic contributions or authors' work in scientific texts

work contribution

No Comments

No Known Activations