INDEX

Explanations

references to legislative or judicial processes

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

proper names followed by verbs

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

 oposición

-0.40

 identitas

-0.40

 trône

-0.38

AccessorTable

-0.36

Mikro

-0.36

 ainfi

-0.35

ambién

-0.35

exitRule

-0.35

 manuales

-0.35

 difesa

-0.35

POSITIVE LOGITS

出版年

0.46

 يتيمه

0.45

ngrx

0.45

 initi

0.45

Epis

0.43

addContainerGap

0.43

EndContext

0.42

 елның

0.42

 oprot

0.42

 bezeichneter

0.42

Activations Density 0.096%

references to legislative or judicial processes

proper names followed by verbs

No Comments

No Known Activations