INDEX

Explanations

questions related to how to perform specific tasks or actions

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

how can i

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

MessageOf

-0.83

 ModelExpression

-0.80

 disambiguazione

-0.77

 queſta

-0.77

ChildScrollView

-0.76

TestingModule

-0.75

principalColumn

-0.74

 виправивши

-0.73

 Мексичка

-0.72

ruik

-0.72

POSITIVE LOGITS

0.40

Bre

0.39

how

0.36

us

0.35

howto

0.34

to

0.34

0.33

Activations Density 0.062%

questions related to how to perform specific tasks or actions

how can i

No Comments

No Known Activations