INDEX

Explanations

occurrences of the term "work out" in various contexts, particularly relating to calculations or evaluations

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

work out

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

It detects prompts asking the reader to perform a mathematical calculation or work out a problem (imperative math-question markers).

oai_token-act-pair · gpt-5-mini Triggered by @vetterc0

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 flavorful

-0.63

 favorably

-0.63

 theater

-0.62

 ardor

-0.61

 favorite

-0.61

 mustache

-0.60

favorite

-0.60

 theaters

-0.60

 favorable

-0.59

 confiable

-0.58

POSITIVE LOGITS

Whilst

1.32

 Whilst

1.31

 whilst

1.22

setVerticalGroup

1.17

initialise

1.16

 initialise

1.15

 Initialise

1.14

 recognisable

1.11

 recognised

1.09

 behavioural

1.09

Activations Density 8.281%

occurrences of the term "work out" in various contexts, particularly relating to calculations or evaluations

work out

It detects prompts asking the reader to perform a mathematical calculation or work out a problem (imperative math-question markers).

No Comments

No Known Activations