INDEX

Explanations

specific nouns and terms related to family and care

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

objects followed by punctuation

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

FECT

-0.35

を中心に

-0.33

 incorrect

-0.33

incorrect

-0.32

usammen

-0.31

angka

-0.31

bieg

-0.30

AM

-0.30

WRONG

-0.30

 rapports

-0.30

POSITIVE LOGITS

 صوتيه

0.76

oredCriteria

0.65

 JpaRepository

0.61

 '\\;'

0.59

 disambiguazione

0.57

 سكانية

0.57

 ویکی‌آمباردا

0.55

 <<<<<<<<<<<<<<

0.55

存于互联网档案馆

0.55

addCriterion

0.55

Activations Density 0.001%

specific nouns and terms related to family and care

objects followed by punctuation

No Comments

No Known Activations