INDEX

Explanations

references to various forms of summoning or calling

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

summon or subpoena

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

tanleria

-0.49

 Infórmanos

-0.46

jenige

-0.45

Хьажоргаш

-0.44

ymce

-0.43

ölkerung

-0.43

fillType

-0.42

 mesmas

-0.42

 laiss

-0.40

 työn

-0.40

POSITIVE LOGITS

 summons

0.91

 Summon

0.89

 summoned

0.88

 summoning

0.85

 summon

0.84

summon

0.84

Summon

0.84

 invitation

0.66

 invitations

0.65

invitation

0.64

Activations Density 0.017%

references to various forms of summoning or calling

summon or subpoena

No Comments

No Known Activations