INDEX

Explanations

instances of accusations or controversies, particularly involving uniforms or authorities

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

suppress the wall

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

 aspect

-0.33

 meaning

-0.31

 exact

-0.30

 ktorá

-0.30

cześ

-0.29

żdy

-0.29

aspetto

-0.28

 entire

-0.28

AndEndTag

-0.28

 localObject

-0.27

POSITIVE LOGITS

ロウィン

0.64

ंदीखरीदारी

0.64

uxxxx

0.60

Nub

0.59

ésultats

0.59

 betweenstory

0.59

istoitu

0.59

ſſung

0.57

RenderAtEndOf

0.57

expandindo

0.56

Activations Density 0.153%

instances of accusations or controversies, particularly involving uniforms or authorities

suppress the wall

No Comments

No Known Activations