INDEX

Explanations

news agency names and associated identifiers

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

AP/ANI location tags

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

 praktischen

-0.46

 englischen

-0.42

demás

-0.41

➯

-0.40

 PROFES

-0.38

 Spitze

-0.38

 LIMITED

-0.37

 Internasional

-0.37

 Jenderal

-0.37

новременно

-0.36

POSITIVE LOGITS

ANSA

0.67

KommentareTeilen

0.60

 Numerade

0.58

 disambiguazione

0.57

 HasFactory

0.56

 ujednoznacz

0.56

Bronnen

0.55

ьаж

0.54

ValueStyle

0.54

complexContent

0.54

Activations Density 0.002%

news agency names and associated identifiers

AP/ANI location tags

No Comments

No Known Activations