INDEX

Explanations

numerical data and patterns involving layered structures or sequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

case

-0.59

Кла

-0.46

 case

-0.46

sez

-0.45

SEG

-0.45

seg

-0.45

def

-0.44

Clas

-0.43

 Case

-0.43

ley

-0.41

POSITIVE LOGITS

 AssemblyCulture

0.85

 صوتيه

0.76

+#+#

0.73

LookAnd

0.71

RegistryLite

0.70

 invokingState

0.70

nologue

0.66

DockStyle

0.66

✭✭

0.66

 AssemblyTitle

0.66

Activations Density 0.630%

numerical data and patterns involving layered structures or sequences

No Comments

No Known Activations