INDEX

Explanations

technical terms related to specific medical or scientific contexts

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

technical descriptions of systems

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

 ogóle

-0.49

?».

-0.49

leſs

-0.47

%</

-0.46



-0.45

Jereo

-0.45

ínű

-0.45

ellido

-0.44

 nemlig

-0.44

ſche

-0.44

POSITIVE LOGITS

 disambiguazione

0.63

complexContent

0.61

withIdentifier

0.47

 autorytatywna

0.45

WebVitals

0.45

enterOuterAlt

0.44

AutoField

0.43

complexType

0.39

PMailer

0.39

 service

0.38

Activations Density 0.033%

technical terms related to specific medical or scientific contexts

technical descriptions of systems

No Comments

No Known Activations