INDEX

Explanations

references to individuals involved in research or interview studies

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

involved in studies or events

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

 Normdatei

-1.10

+#+#

-0.95

IntoConstraints

-0.87

 Мексичка

-0.86

MigrationBuilder

-0.86

HomeAsUpEnabled

-0.84

 SafeMath

-0.83

IVEREF

-0.82

ChildScrollView

-0.82

posedge

-0.81

POSITIVE LOGITS

 hereof

0.59

 thereof

0.52

 directos

0.48

 persoane

0.48

及其

0.47

ámara

0.47

 Personen

0.47

 circonstances

0.46

となる

0.45

 utama

0.44

Activations Density 0.217%

references to individuals involved in research or interview studies

involved in studies or events

No Comments

No Known Activations