INDEX

Explanations

phrases indicating the acceptance of manuscripts for publication

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

accepted for publication

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

tur

-0.36

tit

-0.35

 demon

-0.35

pin

-0.35

emp

-0.34

ؤلاء

-0.34

=("

-0.34

̥

-0.33

 evening

-0.33

 feeling

-0.33

POSITIVE LOGITS

 Accepted

1.27

Accepted

1.23

accepted

1.07

 accepted

1.03

ACCEPTED

0.88

 Acceptance

0.84

 Accept

0.80

 accepte

0.79

Acceptance

0.77

 ACCEPT

0.75

Activations Density 0.002%

phrases indicating the acceptance of manuscripts for publication

accepted for publication

No Comments

No Known Activations