INDEX

Explanations

references to programming concepts and structures, particularly related to types and class definitions

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

Group One characteristics

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

 InputDecoration

-0.63

SBATCH

-0.61



-0.57

 CallOverrides

-0.57

 informée

-0.55

sidemargin

-0.55

Personensuche

-0.54

ſelf

-0.54

 CommonModule

-0.54

didReceive

-0.53

POSITIVE LOGITS

 læng

0.33

 nød

0.32

RTSN

0.32

 moneda

0.31

 AssemblyVersion

0.30

hæng

0.28

 gedrag

0.27

 vervol

0.27

 suficientes

0.27

 tels

0.27

Activations Density 0.769%

references to programming concepts and structures, particularly related to types and class definitions

Group One characteristics

No Comments

No Known Activations