INDEX

Explanations

proper nouns, specifically names of people and places

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

monomeric or assembly completion

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 légère

-0.41

 zato

-0.39

 Angeb

-0.37

 appui

-0.37

 assolu

-0.36

taskId

-0.36

 unknownFields

-0.35

 yayınları

-0.34

iento

-0.34

 ligera

-0.34

POSITIVE LOGITS

SequentialGroup

0.61

 Normdatei

0.59

endpush

0.53



0.52

তথ্যসূত্র

0.49

PerformLayout

0.47

softc

0.47

 AssemblyCompany

0.46

 مرئيه

0.45

ArgumentParser

0.44

Activations Density 0.048%

proper nouns, specifically names of people and places

monomeric or assembly completion

No Comments

No Known Activations