INDEX

Explanations

mathematical and statistical notations or concepts related to assumptions and simplifications in equations

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

special case

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

 يتيمه

-0.77

ⓧ

-0.70

Notae

-0.63

RTDA

-0.62

onViewCreated

-0.62

 Normdatei

-0.60

writeFieldEnd

-0.60

YourGuide

-0.59

 AssemblyCompany

-0.59

 AssemblyCulture

-0.59

POSITIVE LOGITS

0.36

 condition

0.35

Ex

0.34

Pic

0.33

 létre

0.33

can

0.33

 erwarten

0.33

 hence

0.32

 simplicité

0.32

 princí

0.32

Activations Density 0.247%

mathematical and statistical notations or concepts related to assumptions and simplifications in equations

special case

No Comments

No Known Activations