INDEX

Explanations

historical dates or time periods

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

numbers in front of user

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

 doInBackground

-0.83

Бахар

-0.79

 ſind

-0.77

istoitu

-0.77



-0.75

 Cæsar

-0.75

ſchaft

-0.74

iſten

-0.74

 indígen

-0.73

 GetEnumerator

-0.73

POSITIVE LOGITS

ขึ้น

0.26

using

0.24

 Regierung

0.21

一口气

0.21

 administrativos

0.21

 médicos

0.20

 Getreide

0.20

 usos

0.20

 aident

0.20

 Unterrichts

0.20

Activations Density 0.056%

historical dates or time periods

numbers in front of user

No Comments

No Known Activations