INDEX

Explanations

the word "hence" as a signal of causation or reasoning

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

PSO

-0.44

ropical

-0.44

 Travelers

-0.43

 istr

-0.43

AAP

-0.42

updateUser

-0.42

 updateUser

-0.42

 Pullman

-0.42

 "*",

-0.41

tuer

-0.41

POSITIVE LOGITS

hence

0.83

Hence

0.74

 Hence

0.72

 hence

0.71

 defaultstate

0.67

arangay

0.60

mbalikan

0.60

 nadzie

0.57

 Hesse

0.57

/**

0.57

Activations Density 0.011%

the word "hence" as a signal of causation or reasoning

No Comments

No Known Activations