INDEX

Explanations

active verbs related to various actions or processes

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

actions and transactions

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

SequentialGroup

-1.11

ロウィン

-0.94

addPreferredGap

-0.83

 propOrder

-0.83

<unused51>

-0.82

<unused41>

-0.82

 メンテナ

-0.82

<unused8>

-0.82

niſſe

-0.82

<unused14>

-0.81

POSITIVE LOGITS

 this

0.46

0.44

enumi

0.39

 them

0.38

 Ganze

0.33

 conférences

0.32

 ruines

0.32

 biens

0.30

 these

0.29

 trouvez

0.29

Activations Density 0.094%

active verbs related to various actions or processes

actions and transactions

No Comments

No Known Activations