INDEX

Explanations

terms related to decreasing quantities or reductions in various contexts

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

quitous

-0.48

 welder

-0.47

 mixers

-0.47

 sewers

-0.45

 resorts

-0.45

Personendaten

-0.44

Goy

-0.44

Cyfarwyddwr

-0.44

Activator

-0.43

 zippers

-0.43

POSITIVE LOGITS

 reduced

0.78

 reduction

0.75

 decrease

0.74

 diminution

0.70

 decreased

0.69

 reduce

0.68

 reducing

0.67

減

0.66

 Reduced

0.66

Reduced

0.66

Activations Density 0.978%

terms related to decreasing quantities or reductions in various contexts

No Comments

No Known Activations