INDEX

Explanations

references to official titles or positions within organizational contexts

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

appointments and statuses

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

\{\\

-0.65

AddTagHelper

-0.57

setVerticalGroup

-0.57

SqlSession

-0.56

stdc

-0.56

 tartalomajánló

-0.53

ContentAsync

-0.53

GEBURTS

-0.51

ArgsConstructor

-0.51

OGND

-0.50

POSITIVE LOGITS

 Keuangan

0.40

 menetap

0.38

 berätt

0.38

 Kennt

0.36

 proposée

0.36

 publicada

0.35

Tracce

0.35

 Meldung

0.35

 disambiguazione

0.35

MENTE

0.35

Activations Density 0.061%

references to official titles or positions within organizational contexts

appointments and statuses

No Comments

No Known Activations