INDEX

Explanations

references to junctions in various contexts

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

junctions

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

<blockquote>

-0.42

<tr>

-0.41

}))

-0.41

 Públicas

-0.39

<b>

-0.38

IBar

-0.38

korb

-0.37

Ră

-0.37

Steve

-0.36

vânt

-0.36

POSITIVE LOGITS

 Junction

1.32

 junction

1.27

Junction

1.25

 junctions

1.19

junction

1.09

 juncture

0.78

JUN

0.75

 Juniper

0.74

 defaultstate

0.71

 Intersection

0.70

Activations Density 0.005%

references to junctions in various contexts

junctions

No Comments

No Known Activations

references to junctions in various contexts

junctions

No Comments

No Known Activations