INDEX

Explanations

references to historical and geographical entities

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

historical dates and figures

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

tvguidetime

-0.57

 resourceCulture

-0.54

ſelf

-0.53

twimg

-0.51

yves

-0.49

kraine

-0.48

 uſe

-0.47

RenderAtEndOf

-0.47

 vooz

-0.46

fjspx

-0.46

POSITIVE LOGITS

PreferredItem

0.47

RRect

0.42

 zahrani

0.41

 vindo

0.40

 diku

0.40

 informée

0.39

OGND

0.39

 étranger

0.39

apprendre

0.36

Bibliograf

0.36

Activations Density 0.014%

references to historical and geographical entities

historical dates and figures

No Comments

No Known Activations