INDEX

Explanations

evidence and claims related to legal and clinical assessments

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

must not or will need

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

 betweenstory

-1.00

setVerticalGroup

-0.74

 queſta

-0.73

oa̍t

-0.73

TypedDataSet

-0.71

 kohdetta

-0.70

URLException

-0.65

webElementXpaths

-0.65

ніципа

-0.64

SharedDtor

-0.63

POSITIVE LOGITS

 dedans

0.36

 sonde

0.32

WebVitals

0.32

 cinta

0.31

'{@

0.28

 esportivo

0.28

spyOn

0.27

 MediatR

0.27

the

0.27

 fuerte

0.27

Activations Density 0.046%

evidence and claims related to legal and clinical assessments

must not or will need

No Comments

No Known Activations