INDEX

Explanations

specific alphanumeric codes or symbols associated with data entries or statistics

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

medical abbreviations and codes

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

tagHelperRunner

-0.53

 Utilizamos

-0.52

 DISTINCT

-0.46

 neutre

-0.45

 démarche

-0.44

 dentaire

-0.43

DebuggerNonUser

-0.42

 extranjero

-0.41

 suprême

-0.41

 embri

-0.41

POSITIVE LOGITS

Autoritní

0.58

fromnode

0.57

 nakalista

0.57

imageshack

0.53

UserScript

0.52

Билгалдахарш

0.51

 autorytatywna

0.50

 ostavi

0.50

endphp

0.49

 MutableLiveData

0.47

Activations Density 0.255%

specific alphanumeric codes or symbols associated with data entries or statistics

medical abbreviations and codes

No Comments

No Known Activations