INDEX

Explanations

quotes or references that denote names, identities, and roles

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

Single letters and symbols

np_acts-logits-general · gemini-2.0-flash

residue levels, underscores with indices, angle brackets with JS

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 Mean

-0.47

 mean

-0.46

mean

-0.45

gql

-0.42

za

-0.42

 língua

-0.42

DataPropertyName

-0.41

ellees

-0.40

 heraus

-0.40

 geçti

-0.39

POSITIVE LOGITS

 وتسجيلات

0.73

Portale

0.67

########.

0.66

 مرئيه

0.65

IUrlHelper

0.65

ⓧ

0.65

 itſelf

0.64

MessageOf

0.64

 bezeichneter

0.64

 AssemblyCulture

0.61

Activations Density 0.048%

quotes or references that denote names, identities, and roles

Single letters and symbols

residue levels, underscores with indices, angle brackets with JS

No Comments

No Known Activations