INDEX

Explanations

references to beams and their characteristics in an experimental context

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

Nathalie

-0.55

prefixer

-0.53

 Nathalie

-0.51

 Virginie

-0.51

LLocation

-0.50

 étranger

-0.47

Спољашње

-0.46

ทั่ว

-0.46

 rappeler

-0.43

睹

-0.43

POSITIVE LOGITS

Beam

1.27

 Beam

1.23

 beam

1.13

 BEAM

1.09

beam

1.09

 Beams

0.93

BEAM

0.89

 beams

0.84

beams

0.79

Bea

0.76

Activations Density 0.009%

references to beams and their characteristics in an experimental context

No Comments

No Known Activations