INDEX

Explanations

definite articles and pronouns in multiple languages

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

articles in multiple languages

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

 ModelRenderer

-0.38

etje

-0.37

üğ

-0.36

 produisons

-0.35

Numerade

-0.35

 Inscrivez

-0.35

ųjų

-0.34

foglal

-0.34

 feroit

-0.33

TODO

-0.33

POSITIVE LOGITS

 חיצוניים

0.61

riwal

0.58

 gynnwys

0.54

帖最后由

0.54

hilation

0.53

 consiguiente

0.51

何より

0.50

 reciclaje

0.49

sequently

0.49

!*\

0.48

Activations Density 0.002%

definite articles and pronouns in multiple languages

articles in multiple languages

No Comments

No Known Activations