INDEX

Explanations

programming and technical language, particularly related to module definitions and configurations

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

Code snippets related to modules/drivers

np_acts-logits-general · gemini-2.0-flash

code module identifiers

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

PYX

-0.34

 squeeze

-0.31

 πρό

-0.29

squeeze

-0.26

åt

-0.25

 Dias

-0.25

ktı

-0.24

der

-0.24

el

-0.24

 honour

-0.24

POSITIVE LOGITS

 typelib

0.83

 للمعارف

0.77

tagHelperRunner

0.77

rungsseite

0.77

WebServlet

0.76

 パンチラ

0.71

஡

0.71

<unused52>

0.70

<unused47>

0.70

<pad>

0.70

Activations Density 0.361%

programming and technical language, particularly related to module definitions and configurations

Code snippets related to modules/drivers

code module identifiers

No Comments

No Known Activations