INDEX

Explanations

descriptions of size or dimensions

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

size and dimension

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

Derbyniad

-0.57

DataAnnotations

-0.54

takan

-0.51

AntiForgeryToken

-0.47

WebElementEntity

-0.47

AddTagHelper

-0.46

 препратки

-0.46

ksikon

-0.46

 dormitórios

-0.46

uș

-0.44

POSITIVE LOGITS

 themſelves

0.38

ſelves

0.38

 numberOfRows

0.36

GOTREF

0.35

neſs

0.35

 nahilalakip

0.34

 autorytatywna

0.34

Enough

0.34

slidesToShow

0.33

 inte

0.33

Activations Density 0.030%

descriptions of size or dimensions

size and dimension

No Comments

No Known Activations