INDEX

Explanations

mentions of specific references or citations

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

user input turn

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

 Gerv

-0.43

Uncle

-0.42

もちゃ

-0.41

Hozzáférés

-0.40

 aufzu

-0.38

tut

-0.38

 Référence

-0.38

Electro

-0.37

Be

-0.36

 vær

-0.35

POSITIVE LOGITS

出版年

0.72

contentLoaded

0.63

 autorytatywna

0.61

NameInMap

0.59

исленность

0.59

 lenker

0.57

 CreateTagHelper

0.54

üyada

0.52

IUrlHelper

0.51

 pleaſure

0.49

Activations Density 0.000%

mentions of specific references or citations

user input turn

No Comments

No Known Activations