INDEX

Explanations

words and phrases related to legal terms and processes

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

Followed by "is", "are", "was", "would", or "could"

np_acts-logits-general · gemini-2.0-flash

states of being

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.28.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Personendaten

-0.75

 للاسماء

-0.68

setVerticalGroup

-0.68

 ujednoznacz

-0.61

verwijspagina

-0.54

UnitTesting

-0.54

 básicamente

-0.54

 basically

-0.53

 naprawdę

-0.51

resaid

-0.51

POSITIVE LOGITS

 internetowa

0.45

 بخوانید

0.42

+#+

0.39

 huden

0.39

getElementById

0.38

 victor

0.34

tiftung

0.33

 anti

0.33

 Autorizaciones

0.33

 vilja

0.33

Activations Density 1.338%

words and phrases related to legal terms and processes

Followed by "is", "are", "was", "would", or "could"

states of being

No Known Activations