INDEX

Explanations

phrases indicating opinions or evaluations related to benefits and drawbacks

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

good for

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

 persoane

-0.50

 stället

-0.47

 tewas

-0.44

 senere

-0.43

 presented

-0.43

 prisão

-0.43

queles

-0.43

 către

-0.43

 montanha

-0.42

 steder

-0.42

POSITIVE LOGITS

GOOD

0.63

 GOOD

0.63

healthy

0.63

good

0.63

 BoxDecoration

0.63

+#+#

0.62

Healthy

0.59

Good

0.58

hurt

0.57

IUrlHelper

0.57

Activations Density 0.011%

phrases indicating opinions or evaluations related to benefits and drawbacks

good for

No Comments

No Known Activations