INDEX

Explanations

instances of verbosity and readability in programming and textual contexts

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

 iPads

-0.49

Shat

-0.48

 pegat

-0.47

 Shoulders

-0.47

://$

-0.47

Deal

-0.47

 Parton

-0.47

Chicken

-0.46

 poule

-0.46

KommentareTeilen

-0.46

POSITIVE LOGITS

verbose

1.17

 verbose

1.05

Verbose

1.03

VERBOSE

0.93

 MessageBoxIcon

0.61

bosity

0.58

bose

0.57

arangay

0.54

 VERB

0.54

setTextSize

0.52

Activations Density 0.002%

instances of verbosity and readability in programming and textual contexts

No Comments

No Known Activations