INDEX

Explanations

technical terms and statistical language related to scientific analysis

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

Changes in numerical values

np_acts-logits-general · gemini-2.0-flash

observed decreases or trends

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

enumi

-0.59

 Endless

-0.54

ügen

-0.52

 đương

-0.52

GraphicsUnit

-0.52

mybatisplus

-0.51

Geplaatst

-0.50

Endless

-0.49

 незавершена

-0.48

 lebens

-0.48

POSITIVE LOGITS

 values

0.99

 valores

0.88

 observed

0.88

 slightly

0.87

 decreases

0.86

 indicating

0.83

 decrease

0.83

 trend

0.81

 valori

0.80

 valeurs

0.80

Activations Density 2.390%

technical terms and statistical language related to scientific analysis

Changes in numerical values

observed decreases or trends

No Comments

No Known Activations