INDEX

Explanations

terms and phrases related to relevance

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

relevance

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

Gou

-0.41

 ház

-0.40

Machine

-0.39

Gou

-0.39

Physician

-0.39

shoes

-0.39

shoe

-0.38

 Shoemaker

-0.38

BarStyle

-0.37

 house

-0.37

POSITIVE LOGITS

 relevance

1.07

 Relevance

0.98

 Relevant

0.96

Relevant

0.96

 relevant

0.93

relevant

0.92

Relevance

0.91

relevance

0.90

 Irrelevant

0.81

 relevan

0.79

Activations Density 0.013%

terms and phrases related to relevance

relevance

No Comments

No Known Activations