INDEX

Explanations

occurrences of special identifiers and modular structures in code

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

starts of turns or code blocks

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

rungsseite

-0.63

 otomatig

-0.60

liferay

-0.59

 Chwiliwch

-0.59

erai

-0.58

rogels

-0.57

errHandler

-0.57

Rhestr

-0.56

ymce

-0.56

хьтан

-0.56

POSITIVE LOGITS

'__

0.74

"__

0.67

✭✭

0.44

"__

0.42

$__

0.42

 Batal

0.39

("__

0.38

 Houſe

0.37

ſelves

0.37

дмила

0.37

Activations Density 0.001%

occurrences of special identifiers and modular structures in code

starts of turns or code blocks

No Comments

No Known Activations