INDEX

Explanations

references to object-oriented programming concepts

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

null or object

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

ngdoc

-0.58

сылкі

-0.56

 Biôgrafia

-0.55

SuppressLint

-0.52

RTEE

-0.51

Hood

-0.48

 jumper

-0.48

 Minor

-0.47

 Hood

-0.47

endpush

-0.47

POSITIVE LOGITS

Personendaten

0.66

 literaria

0.45

 ludzi

0.43

 zijne

0.43

StoryboardSegue

0.42

 henkil

0.42

 noiva

0.41

 pihaknya

0.41

 enige

0.40

(!__

0.40

Activations Density 0.005%

references to object-oriented programming concepts

null or object

No Comments

No Known Activations