INDEX

Explanations

phrases indicating location and directionality

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

Following the word "from" and locations

np_acts-logits-general · gemini-2.0-flash

locations and directions

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

SBATCH

-0.59

cshtml

-0.59

doms

-0.53

CWE

-0.52

ecycle

-0.51

TagMode

-0.50

tagHelperRunner

-0.50

createStatement

-0.49

#+#

-0.49

ftagPool

-0.49

POSITIVE LOGITS

 location

0.68

 Location

0.64

 directions

0.63

 parking

0.60

 locating

0.60

 locate

0.58

 распо

0.56

location

0.56

 LOCATION

0.56

directions

0.55

Activations Density 0.139%

phrases indicating location and directionality

Following the word "from" and locations

locations and directions

No Comments

No Known Activations