INDEX

Explanations

references to theorems and mathematical statements

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

theorem labels

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 незавершена

-0.75

########.

-0.60

transQ

-0.57

InjectAttribute

-0.57

PreferredItem

-0.56

RegressionTest

-0.56

AnchorStyles

-0.55

(!__

-0.55

 мәкал

-0.55

IsMutable

-0.54

POSITIVE LOGITS

 GenerationType

0.41

 Dickinson

0.39

GCT

0.36

 LoginPage

0.36

匿名使用者

0.34

fromnode

0.34

chapp

0.34

 Anſ

0.34

 Serbie

0.33

ISE

0.33

Activations Density 0.002%

references to theorems and mathematical statements

theorem labels

No Comments

No Known Activations