INDEX

Explanations

phrases related to justification or reasons for actions and conditions

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

OrBuilder

-0.51

 Insee

-0.43

BeginInit

-0.41

 vermelhas

-0.40

ⓧ

-0.39

 fieldNum

-0.37

 AttributeSet

-0.36

Formal

-0.35

JSP

-0.35

 Sides

-0.35

POSITIVE LOGITS

 wegen

0.59

 akibat

0.56

 aufgrund

0.54

ibatkan

0.52

httphttps

0.52

 Aufgrund

0.52

 devido

0.51

 Gründe

0.51

SequentialGroup

0.50

 dėl

0.50

Activations Density 0.082%

phrases related to justification or reasons for actions and conditions

No Comments

No Known Activations