INDEX

Explanations

mentions of computational concepts and terminology

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

computational fields and methods

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

 Inscrivez

-0.69

rawDesc

-0.60

 Bewußt

-0.55

 Garantía

-0.54

 ainfi

-0.54

 suivante

-0.54

 voici

-0.53

 vermelhas

-0.53

 Glied

-0.52

reszcie

-0.52

POSITIVE LOGITS

 computational

0.90

 Computational

0.77

Computational

0.75

neurial

0.65

ational

0.57

ATIONAL

0.56

perative

0.53

 Compton

0.52

Alg

0.52

lational

0.51

Activations Density 0.005%

mentions of computational concepts and terminology

computational fields and methods

No Comments

No Known Activations