INDEX

Explanations

database-related terms, specifically focusing on keys in table structures

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

mass redistribution

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

-0.68

parsedMessage

-0.64

LabelTagHelper

-0.62

IsMutable

-0.60

aarrggbb

-0.57

艋

-0.55

lapsingToolbar

-0.55

Precautionary

-0.54

OrNil

-0.54

XmlEnum

-0.52

POSITIVE LOGITS

PrimaryKey

0.52

 EconPapers

0.48

HasKey

0.47

rawQuery

0.38

 Chwiliwch

0.36

 htons

0.35

Autoritní

0.34

setViewportView

0.33

Kaynakça

0.30

 independent

0.30

Activations Density 0.000%

database-related terms, specifically focusing on keys in table structures

mass redistribution

No Comments

No Known Activations