INDEX

Explanations

references to significant institutions or organizations

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

organizations and locations

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

rrggbb

-0.63

wieś

-0.57

 almofada

-0.53

 sentarse

-0.53

Personensuche

-0.50

 visualización

-0.50

 cadeira

-0.48

 fotografico

-0.48

findpost

-0.48

ValueGeneration

-0.46

POSITIVE LOGITS

StructEnd

0.52

CppCodeGen

0.43

MockMvc

0.41

Esa

0.38

CU

0.38

 مشين

0.37

Bend

0.36

drawRect

0.36

 Scalars

0.35

RestTemplate

0.35

Activations Density 0.092%

references to significant institutions or organizations

organizations and locations

No Comments

No Known Activations