INDEX

Explanations

graph-related syntax and structures

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ویکی‌پدی

-1.00

Personendaten

-0.96

 typelib

-0.88

 kasarigan

-0.84

contentLoaded

-0.82

GEBURTSDATUM

-0.81

 betweenstory

-0.80

Autoritní

-0.79

transQ

-0.78

featureID

-0.77

POSITIVE LOGITS

arr

0.36

гон

0.35

 mark

0.35

Pin

0.35

gan

0.35

rac

0.34

 degree

0.34

mas

0.34

Sera

0.34

Activations Density 0.636%

graph-related syntax and structures

No Comments

No Known Activations