INDEX

Explanations

references to authors and their affiliations in scientific literature

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

academic phrases and institutions

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

 betweenstory

-0.72

Personendaten

-0.65

verwijspagina

-0.62

PYX

-0.60

__':

-0.59

enderror

-0.57

новништво

-0.56

ValueStyle

-0.55

 Autorizaciones

-0.55

 surla

-0.55

POSITIVE LOGITS

<eos>

0.40

 kaldır

0.36

的过程中

0.35

appan

0.33

 bancaria

0.32

🏻

0.32

rog

0.32

 Masse

0.31

 Chriftian

0.31

vs

0.31

Activations Density 0.023%

references to authors and their affiliations in scientific literature

academic phrases and institutions

No Comments

No Known Activations