INDEX

Explanations

terms related to blades and knives

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ftagPool

-0.55

 Komunikasi

-0.52

transQ

-0.52

isDebugEnabled

-0.49

 Gestão

-0.48

Datuak

-0.46

 Taktlose

-0.46

órmula

-0.45

XtraReports

-0.44

 Keuangan

-0.44

POSITIVE LOGITS

 knife

0.60

 knives

0.58

 Knife

0.54

 Knives

0.51

Knife

0.48

knife

0.45

 blade

0.41

 cuchillo

0.40

 Blade

0.38

EDC

0.38

Activations Density 0.002%

terms related to blades and knives

No Comments

No Known Activations