INDEX

Explanations

phrases indicating relationships or connections, particularly involving the term "of."

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

qualities of soundness

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

addPreferredGap

-0.68

 EconPapers

-0.56

flashdata

-0.54

ništ

-0.52

AsUp

-0.52

principalColumn

-0.52

 يتيمه

-0.49

expandindo

-0.48

cerpt

-0.48

bleau

-0.47

POSITIVE LOGITS

sibility

0.45

 isComment

0.43

genicity

0.43

acidade

0.41

 soundness

0.40

Trust

0.40

 bambú

0.40

 trust

0.40

silience

0.39

вости

0.39

Activations Density 0.069%

phrases indicating relationships or connections, particularly involving the term "of."

qualities of soundness

No Comments

No Known Activations