INDEX

Explanations

references to trees and tree-related imagery

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

trees

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

AddTagHelper

-0.94

 EconPapers

-0.84

 InputDecoration

-0.84

setVerticalGroup

-0.79

AndEndTag

-0.74

 SwitchCompat

-0.72

 surla

-0.70

 kaarangay

-0.69

ScopeManager

-0.68

:]:

-0.66

POSITIVE LOGITS

 trees

0.64

 Trees

0.52

Trees

0.42

 TREES

0.42

 tree

0.41

 ağaç

0.41

 Bäume

0.40

 árboles

0.40

 arbres

0.37

 деревья

0.37

Activations Density 0.003%

references to trees and tree-related imagery

trees

No Comments

No Known Activations