INDEX

Explanations

specific terms related to technical and operational aspects of software or programming

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

bogue, releasably, wantonly vile, Perdot, tortious injury

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

Autoritní

-0.62

parsedMessage

-0.60

########.

-0.56

 betweenstory

-0.47

NameInMap

-0.44

AutoScaleMode

-0.43

évaluateur

-0.42

 informée

-0.41

queryInterface

-0.41

forName

-0.40

POSITIVE LOGITS

ghijklmnop

0.58

alakip

0.50

arangay

0.49

AddTagHelper

0.48

 ſur

0.47

 avoient

0.46

tiérrez

0.45

 Chhat

0.44

GeoNames

0.44

 étoient

0.44

Activations Density 0.028%

specific terms related to technical and operational aspects of software or programming

bogue, releasably, wantonly vile, Perdot, tortious injury

No Comments

No Known Activations