INDEX

Explanations

document structure or formatting indicators

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

code snippets and legal document headers

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

 exposición

-0.35

highlighted

-0.33

 serons

-0.33

WebElementEntity

-0.33

priest

-0.32

 priest

-0.32

 highlighted

-0.32

Théâtre

-0.31

 suspensión

-0.31

 troup

-0.30

POSITIVE LOGITS

RenderAtEndOf

0.59

 pinulongan

0.58

 utafitiHapana

0.58

AndEndTag

0.54

WriteTagHelper

0.54

 <=",

0.52

ButterKnife

0.50



0.50

LayoutStyle

0.49

 iconLine

0.48

Activations Density 0.123%

document structure or formatting indicators

code snippets and legal document headers

No Comments

No Known Activations