INDEX

Explanations

structured data and specific identifiers within programming or database contexts

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

mathematical notation or programming code

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

Datuak

-0.62

ंदीखरीदारी

-0.59

migrationBuilder

-0.42

发表于

-0.42

oredCriteria

-0.41

BnF

-0.41

EndContext

-0.40

BeginContext

-0.40

phrag

-0.40

おすすめです

-0.38

POSITIVE LOGITS

expandindo

0.54

 المعيارى

0.45

 становника

0.43

verwijspagina

0.41

 mitte

0.41

 beginnetje

0.40

addPreferredGap

0.40

Filmographie

0.40

 通販

0.39

✢

0.38

Activations Density 0.095%

structured data and specific identifiers within programming or database contexts

mathematical notation or programming code

No Comments

No Known Activations