INDEX

Explanations

elements related to user interaction and rights

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

talks to people

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

 nahilalakip

-0.83

MessageTagHelper

-0.77

enderror

-0.72

 CreateTagHelper

-0.71

 CURIAM

-0.70

IBOutlet

-0.69

sizeCache

-0.66

IntoConstraints

-0.66

الحياه

-0.65

 Conſ

-0.65

POSITIVE LOGITS

 calm

0.27

cadi

0.26

Bar

0.25

 probing

0.25

 pesquisa

0.25

értelmű

0.25

 bilang

0.25

 mér

0.24

Activations Density 0.060%

elements related to user interaction and rights

talks to people

No Comments

No Known Activations