INDEX

Explanations

technical specifications and details related to computer architecture or systems

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

numbers followed by unitsize

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

一丝

-0.31

 stab

-0.29

 divul

-0.28

beforeAll

-0.28

cal

-0.28

 larges

-0.27

CELLANEOUS

-0.27

LoadMore

-0.27

شب

-0.27

nut

-0.27

POSITIVE LOGITS

fjspx

0.69

Smaller

0.67

 smaller

0.67

 <<<<<<<<<<<<<<

0.64

smaller

0.64

httphttps

0.63

kháu

0.63

 Smaller

0.61

 kleineren

0.61

RUnlock

0.60

Activations Density 0.098%

technical specifications and details related to computer architecture or systems

numbers followed by unitsize

No Comments

No Known Activations