INDEX

Explanations

terms related to contamination and infection in a medical or environmental context

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

infected or contaminated state

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

 '\\;'

-0.48

Hochspringen

-0.46

verwijspagina

-0.44

(!__

-0.44

 समीक्षाओं

-0.39

 unknownFields

-0.38

 JpaRepository

-0.37

 հղումներ

-0.37

vų

-0.36

 vnto

-0.35

POSITIVE LOGITS

 contaminated

0.71

 infested

0.60

tagHelperRunner

0.60

contaminated

0.59

 infected

0.58

Filled

0.53

Personendaten

0.51

fromnode

0.50

infected

0.50

 afflicted

0.49

Activations Density 0.010%

terms related to contamination and infection in a medical or environmental context

infected or contaminated state

No Comments

No Known Activations