INDEX

Explanations

punctuation marks, specifically periods

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-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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tagext

-0.68

WaitGroup

-0.52

transQ

-0.52

SharedCtor

-0.47

 ujednoznacz

-0.46

 disambiguazione

-0.46

abestanden

-0.43

كويكب

-0.42

DockStyle

-0.41

EndContext

-0.41

POSITIVE LOGITS

plotlib

0.62

lammatory

0.57

 تضيفلها

0.52

ocyto

0.51

Wim

0.48

niss

0.47

pylene

0.47

dormit

0.47

ductory

0.46

sertation

0.46

Activations Density 0.000%

punctuation marks, specifically periods

No Comments

No Known Activations