INDEX

Explanations

legal terms and phrases related to certifications and appeals

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

no competing interests

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

Negative Logits

 surla

-0.65

MessageState

-0.49

+#+#

-0.48

 argue

-0.48

 autorytatywna

-0.48

Wea

-0.47

 يتيمه

-0.47

TBI

-0.46

PMI

-0.46

BeginContext

-0.46

POSITIVE LOGITS

AutoScaleMode

0.47

httphttps

0.45

ContextHolder

0.38

 bluzka

0.36

 représent

0.36

AndEndTag

0.35

 represents

0.35

 savu

0.35

EFIAPI

0.34

 skóra

0.33

Activations Density 0.034%

legal terms and phrases related to certifications and appeals

no competing interests

No Known Activations