INDEX

Explanations

phrases related to chemical or scientific measurements and formulations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

often conditional statements

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

Negative Logits

 betweenstory

-0.69

 ujednoznacz

-0.61

請繼續往下閱讀

-0.60

RetentionPolicy

-0.60

contentLoaded

-0.59

queles

-0.57

enablog

-0.56

 sereia

-0.56

kloped

-0.56

 समीक्षक

-0.56

POSITIVE LOGITS

 unless

0.45

 usually

0.41

Rüyada

0.38

unless

0.36

usually

0.36

typically

0.35

 typically

0.33

or

0.33

 except

0.32

 either

0.32

Activations Density 0.214%

phrases related to chemical or scientific measurements and formulations

often conditional statements

No Known Activations