INDEX

Explanations

references to various types of data and their relationships in structured contexts

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

 financing

-0.35

 novel

-0.34

 обо

-0.31

ฐ

-0.31

ぞ

-0.31

掏

-0.31

 ordering

-0.30

 Information

-0.30

 compi

-0.30

 switch

-0.29

POSITIVE LOGITS

 autorytatywna

0.74

 المعيارى

0.71

 <<<<<<<<<<<<<<

0.71

parsedMessage

0.68

MessageTagHelper

0.68

tagHelperRunner

0.67

EDEFAULT

0.66

awtextra

0.65

 مشين

0.59

posedge

0.59

Activations Density 0.085%

references to various types of data and their relationships in structured contexts

No Comments

No Known Activations