INDEX

Explanations

references to specific characters and entities from a franchise or series

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

Meteor breaks

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

 spese

-0.38

PERATURE

-0.37

LookAnd

-0.33

ètres

-0.32

 चीज़ों

-0.32

省市镇

-0.32

ToCheck

-0.32

 Józef

-0.32

 FetchType

-0.31

dff

-0.31

POSITIVE LOGITS

TypedDataSet

0.68

 AssemblyCulture

0.59

 defaultstate

0.56

 Scaling

0.54

DoubleQuotes

0.52

]")]

0.52

abestanden

0.51

XtraReports

0.49

󠁣

0.48

✭✭

0.47

Activations Density 0.018%

references to specific characters and entities from a franchise or series

Meteor breaks

No Comments

No Known Activations