INDEX

Explanations

references to layout orientations and positioning in a visual or spatial context

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

right positioning or delimiters

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

CodeAttribute

-0.75

########.

-0.65

 AssemblyCompany

-0.60

TypedDataSet

-0.55

AddTagHelper

-0.55

modelBuilder

-0.54

LookAnd

-0.54

<unused74>

-0.52

<unused41>

-0.52

[@BOS@]

-0.52

POSITIVE LOGITS

 oldal

0.37

 internetowa

0.37

 ivelany

0.37

 begge

0.37

ViewSet

0.36

]=>

0.34

Népesség

0.33

 NSCoder

0.33

 volna

0.30

üli

0.30

Activations Density 0.002%

references to layout orientations and positioning in a visual or spatial context

right positioning or delimiters

No Comments

No Known Activations