INDEX

Explanations

repeated references to the word "this."

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

questions starting with "is this"

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

 surla

-0.49

mybatisplus

-0.46

 particulières

-0.40

 vertrou

-0.40

 délic

-0.40

 cavalli

-0.40

autonomie

-0.39

heria

-0.39

 rigue

-0.39

httphttps

-0.38

POSITIVE LOGITS

is

0.50

 blitt

0.50

not

0.48

 blevet

0.47

これで

0.47

Hozzáférés

0.47

 เป็น

0.46

 Esto

0.45

 Ayers

0.45

 Это

0.45

Activations Density 0.011%

repeated references to the word "this."

questions starting with "is this"

No Comments

No Known Activations