INDEX

Explanations

negations or denials in the text

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

sentence beginnings

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

 ſche

-0.60

 ویکی‌آمباردا

-0.59

 ſtand

-0.58

 purpoſe

-0.58

ésultats

-0.57

awtextra

-0.57

nexpected

-0.56

 ſind

-0.55

 ſhould

-0.55

ſelves

-0.54

POSITIVE LOGITS

Personendaten

0.33

EDEFAULT

0.33

spraak

0.33

author

0.32

 authors

0.32

Authors

0.32

XtraEditors

0.31

Da

0.29

Fe

0.29

chi

0.28

Activations Density 0.049%

negations or denials in the text

sentence beginnings

No Comments

No Known Activations