INDEX

Explanations

authoritative and positive descriptions related to businesses and institutions

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

wide variety, highly efficient, unique feature

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

GEBURTSDATUM

-0.42

 nakalista

-0.40

 massively

-0.36

 Harus

-0.36

 Major

-0.35

*.

-0.35

uchung

-0.35

sit

-0.34

Cod

-0.34

 Advertisement

-0.34

POSITIVE LOGITS

LookAnd

0.71

RegistryLite

0.65

 كومونز

0.63



0.60

webElementGuid

0.59

ceğine

0.58

DotNetBar

0.57

InputBorder

0.51

 فريبيس

0.51

astify

0.50

Activations Density 0.067%

authoritative and positive descriptions related to businesses and institutions

wide variety, highly efficient, unique feature

No Comments

No Known Activations