INDEX

Explanations

phrases related to scientific proposals and mechanisms

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

scientific and generalization terms

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

 natomiast

-0.45

 lisäksi

-0.34

 murni

-0.34

 bowiem

-0.32

 diejenigen

-0.30

 Königin

-0.29

 pimpinan

-0.29

 semula

-0.28

 directement

-0.27

jenigen

-0.27

POSITIVE LOGITS

хьтан

0.83

ſicht

0.82

iſchen

0.81

 ſeines

0.81

 zwiſchen

0.80

ſelben

0.79

iſche

0.79

 deſſen

0.79

NameInMap

0.78

 daysTop

0.78

Activations Density 0.035%

phrases related to scientific proposals and mechanisms

scientific and generalization terms

No Comments

No Known Activations