INDEX

Explanations

punctuation and numerical patterns in textual content

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

code line numbers

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

<eos>

-0.45

########.

-0.39

ItemBackground

-0.35

 Schilling

-0.31

 shall

-0.29

py

-0.29

 prenota

-0.29

 reservado

-0.28

 Mura

-0.28

 preferably

-0.28

POSITIVE LOGITS

 JADX

1.20

httphttps

0.99

 jadx

0.88

AndEndTag

0.82

 autorytatywna

0.75

 snippetHide

0.71

tagHelperRunner

0.71

帖最后由

0.67

SequentialGroup

0.67

ValueStyle

0.66

Activations Density 0.006%

punctuation and numerical patterns in textual content

code line numbers

No Comments

No Known Activations