INDEX

Explanations

questions, commands, or prompts related to user interaction

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

 disambiguazione

-0.61

 acuarela

-0.40

jsxFileName

-0.38

 فريبيس

-0.38

 ujednoznacz

-0.37

 intereses

-0.36

 cottura

-0.34

 provecho

-0.32

期刊论文

-0.32

ówczas

-0.32

POSITIVE LOGITS

RegressionTest

0.68

SourceChecksum

0.65

parsedMessage

0.64



0.57

skjaer

0.54

TAN

0.53

tanooga

0.47

endam

0.47

IsContent

0.47

sticks

0.46

Activations Density 0.016%

questions, commands, or prompts related to user interaction

No Comments

No Known Activations