INDEX

Explanations

phrases related to ambiguity or uncertainty in statements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ConstraintMaker

-0.56

 Perſ

-0.53

 Anſ

-0.52

 Conſ

-0.51

 Inſ

-0.50

 ſou

-0.48

 Reſ

-0.48

 Diſ

-0.47

 Pristup

-0.47

RegressionTest

-0.46

POSITIVE LOGITS

 justified

0.51

に見

0.39

KommentareTeilen

0.39

 AssemblyCulture

0.39

 AppCompatTheme

0.38

 justifiable

0.38

 gius

0.38

 EconPapers

0.38

༘

0.37

leg

0.37

Activations Density 0.158%

phrases related to ambiguity or uncertainty in statements

No Comments

No Known Activations