INDEX

Explanations

references to olive oil and olive-related terminology

oai_token-act-pair · gpt-4o-mini Triggered by @bot

olive and olives

np_acts-logits-general · gemini-2.5-flash-lite

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

 amounts

-0.42

 Eintritt

-0.40

require

-0.40

 principal

-0.38

faßt

-0.38

yorlar

-0.37

 Barrera

-0.37

 behov

-0.37

Mereka

-0.36

NewReader

-0.36

POSITIVE LOGITS

Olive

1.35

 Olive

1.31

 olive

1.24

olive

1.17

 OLIVE

1.06

 Olives

1.02

 olives

0.97

榄

0.89

 oliva

0.88

 Oliven

0.85

Activations Density 0.005%

references to olive oil and olive-related terminology

olive and olives

No Comments

No Known Activations