INDEX

Explanations

HTML tags and formatting cues

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

HTML and formatting

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



-0.75

 queſta

-0.73

 otomatig

-0.72

sizeCache

-0.72

WebElementEntity

-0.71

########.

-0.71

RenderAtEndOf

-0.67

 TestBed

-0.66

$_(

-0.65

 <>",

-0.64

POSITIVE LOGITS

<strong>

0.42

PO

0.38

0.37

0.36

 Brandt

0.36

centerline

0.35

link

0.35

<em>

0.34

Pres

0.34

Contact

0.34

Activations Density 0.002%

HTML tags and formatting cues

HTML and formatting

No Comments

No Known Activations