INDEX

Explanations

programming-related terms and structures

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

Preceding potential email addresses or usernames

np_acts-logits-general · gemini-2.0-flash

filenames, codes, and email addresses

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 poverty

-0.44

 rodilla

-0.41

 rendah

-0.41

 matters

-0.39

 provisoire

-0.39

 matter

-0.39

ături

-0.38

affaires

-0.38

 katkı

-0.38

low

-0.38

POSITIVE LOGITS

 "..\..\..\

0.64

 "..\..\

0.56

 &___

0.55

TypedDataSet

0.55

 LastName

0.54

 ExecuteAsync

0.54

WriteBarrier

0.52

 AllAfrica

0.51

apimachinery

0.49

AllAfrica

0.49

Activations Density 1.310%

programming-related terms and structures

Preceding potential email addresses or usernames

filenames, codes, and email addresses

No Comments

No Known Activations