INDEX

Explanations

occurrences of the word "the"

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

articles following "the"

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

ofern

-0.35

 Ganzen

-0.32

impanan

-0.31

him

-0.29

 censiti

-0.28

 Landwir

-0.28

 chantier

-0.27

urnya

-0.27

 anniversaire

-0.27

 limpi

-0.27

POSITIVE LOGITS

 EconPapers

0.73

encodeWith

0.62

KommentareTeilen

0.62

protoimpl

0.61

XmlAccessType

0.60

 kasarigan

0.60

 myſelf

0.59

 ſch

0.58

 juſ

0.57

 Administrativna

0.57

Activations Density 0.042%

occurrences of the word "the"

articles following "the"

No Comments

No Known Activations