INDEX

Explanations

specific programming or technical terms related to functions, nodes, and data structures

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

Followed by property, box, node, or tag

np_acts-logits-general · gemini-2.0-flash

node to value or string

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 considérable

-0.38

 remarqué

-0.34

ikis

-0.34

 rarement

-0.34

Tw

-0.34

 McQu

-0.33

optimis

-0.32

 serangan

-0.32

setopt

-0.32

 Schle

-0.32

POSITIVE LOGITS

bootstrapcdn

0.66

 resourceCulture

0.62

 ProtoMessage

0.59

IUrlHelper

0.57

 ब्रेकडाउन

0.57



0.55

 Roskov

0.54

EndContext

0.54

endphp

0.52

eseorang

0.51

Activations Density 0.315%

specific programming or technical terms related to functions, nodes, and data structures

Followed by property, box, node, or tag

node to value or string

No Comments

No Known Activations