INDEX

Explanations

phrases related to conditional reasoning and hypothesis testing

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

"for" or "in" preceding mathematical/logical expressions

np_acts-logits-general · gemini-2.0-flash

mathematical or scientific notation

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

-0.37

ol

-0.33

 eben

-0.31

 mati

-0.31

Con

-0.29

 bottes

-0.29

 lequel

-0.28

POSITIVE LOGITS

 autorytatywna

0.81

 चीज़ों

0.78

rungsseite

0.74

 المعيارى

0.74

pushFollow

0.71

évaluateur

0.66

addPreferredGap

0.65

SequentialGroup

0.65

 incrí

0.65

makeText

0.65

Activations Density 0.620%

phrases related to conditional reasoning and hypothesis testing

"for" or "in" preceding mathematical/logical expressions

mathematical or scientific notation

No Comments

No Known Activations