INDEX

Explanations

specific categories and counts of various entities or objects

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

files, folders, disks, planets

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

 EconPapers

-0.52

󠁴

-0.50

+#+

-0.48

 disambiguazione

-0.46

ագրություններ

-0.45

Soorten

-0.45

 transfieras

-0.44

 Davide

-0.43

zow

-0.43

 hikariConfig

-0.42

POSITIVE LOGITS

 صوتيه

0.43

setAccessible

0.41

ChildScrollView

0.40

jsxFileName

0.39

cèse

0.38

ItemLayout

0.37

AutoScaleMode

0.36

Kjelder

0.35

)}{(

0.35

も多い

0.32

Activations Density 0.260%

specific categories and counts of various entities or objects

files, folders, disks, planets

No Comments

No Known Activations