INDEX

Explanations

references to specific academic institutions and their associated publications or components

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

names, codes, and technical terms

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

transQ

-0.66

 autorytatywna

-0.64

tvguidetime

-0.60

 betweenstory

-0.59

 queſta

-0.59

UserScript

-0.57

Tembelea

-0.57

 الحره

-0.56

 CreateTagHelper

-0.54

AsUp

-0.54

POSITIVE LOGITS

↵↵

0.39

0.38

↵

0.34

Gaw

0.33

 after

0.32

 builders

0.31

Gaw

0.31

 Respon

0.30

.,

0.30

 excluding

0.30

Activations Density 0.010%

references to specific academic institutions and their associated publications or components

names, codes, and technical terms

No Comments

No Known Activations