INDEX

Explanations

sections of text related to announcements and recognitions

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

awards and achievements

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

@"/

-0.45

cute

-0.41

 Köln

-0.41

FormTagHelper

-0.41

bekistan

-0.40

 <<<<<<<<<<<<<<

-0.40

 Akron

-0.40

inge

-0.40

ESR

-0.39

zám

-0.39

POSITIVE LOGITS

 orgull

0.54

 proud

0.54

 Celebrations

0.52

 Congratulations

0.49

 pride

0.49

 congrat

0.48

Congratulations

0.47

 celebra

0.46

 célébr

0.46

 achievement

0.46

Activations Density 0.024%

sections of text related to announcements and recognitions

awards and achievements

No Comments

No Known Activations