INDEX

Explanations

prepositions and phrases indicating purpose or intent

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

relation | attribute | specific noun

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

MigrationBuilder

-0.52

 poveznice

-0.49

oscopy

-0.49

escence

-0.48

Zeneca

-0.48

RefNanny

-0.48

yarnpkg

-0.48

mijn

-0.48

udsman

-0.47

uminense

-0.47

POSITIVE LOGITS

 lateinit

0.42

是从

0.40

 totally

0.38

值为

0.38

 跳转至

0.38

 autorytatywna

0.37

 pertains

0.36

都是在

0.35

internalType

0.35

 proportional

0.35

Activations Density 0.059%

prepositions and phrases indicating purpose or intent

relation | attribute | specific noun

No Comments

No Known Activations