INDEX

Explanations

terminology related to technical processes and components

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

roles, materials, medical terms

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

msgTypes

-0.78

 المعيارى

-0.76

Personendaten

-0.75

Rüyada

-0.74

 مرئيه

-0.73

OGND

-0.71

RegressionTest

-0.71

 Roskov

-0.69

TagMode

-0.69

 betweenstory

-0.69

POSITIVE LOGITS

 respectively

0.36

 break

0.32

nt

0.30

na

0.30

 themselves

0.30

ny

0.29

 both

0.29

 status

0.29

 statue

0.28

Activations Density 0.076%

terminology related to technical processes and components

roles, materials, medical terms

No Comments

No Known Activations