INDEX

Explanations

references to experimental design and optimization processes

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

experiments, optimization, and testing

np_acts-logits-general · gemini-2.0-flash

experimental optimization and testing

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 purpoſe

-0.51

 reaſon

-0.43

cicle

-0.41

casus

-0.41

 Houſe

-0.40

omock

-0.40

 houſe

-0.39

zeh

-0.39

ſehen

-0.38

DrawerToggle

-0.38

POSITIVE LOGITS

 experimentation

0.73

 experimenting

0.71

 AssemblyCulture

0.62

 experiments

0.60

 experimented

0.59

 experimento

0.57

 experiment

0.56

 experimentar

0.53

 experim

0.53

Experiment

0.51

Activations Density 0.841%

references to experimental design and optimization processes

experiments, optimization, and testing

experimental optimization and testing

No Comments

No Known Activations