INDEX

Explanations

terms related to filtering and selection processes

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

DidLoad

-0.52

 NSCoder

-0.50

 Cumber

-0.45

 installa

-0.43



-0.41

 Colleg

-0.41

 installing

-0.40

Carly

-0.40

хьтан

-0.40

sæ

-0.40

POSITIVE LOGITS

 filter

0.83

 filtering

0.76

 filtered

0.72

 lọc

0.71

 filters

0.70

 FILTER

0.68

filter

0.67

 Filtering

0.67

 selectivity

0.67

筛选

0.66

Activations Density 0.078%

terms related to filtering and selection processes

No Comments

No Known Activations