INDEX

Explanations

temporal relationships and time-series data features

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

 referenties

-0.43

 botões

-0.39

richTextPanel

-0.39

 damskie

-0.38

httphttps

-0.36

 alunos

-0.36

 digitais

-0.35

 publicidade

-0.35

 cães

-0.35

ldc

-0.34

POSITIVE LOGITS

 duration

0.68

 duran

0.57

duration

0.57

 viewDidLoad

0.56

dur

0.55

 Duration

0.54

DUR

0.52

Duration

0.52

WriteLiteral

0.51

 durations

0.50

Activations Density 0.170%

temporal relationships and time-series data features

No Comments

No Known Activations