INDEX

Explanations

programming and database-related terminology, particularly focusing on operations involving data manipulation and persistence

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

save, update, return

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

AddTagHelper

-0.52

 disambiguazione

-0.52

tvguidetime

-0.52

 ModelExpression

-0.52

ostoma

-0.51

 jsPsych

-0.50

fromnode

-0.49

SequentialGroup

-0.49

TargetException

-0.49

toare

-0.48

POSITIVE LOGITS

 updated

0.42

更新

0.41

 rebuild

0.41

 rebuilding

0.40

 update

0.40

重新

0.38

 rebuilt

0.38

 updating

0.38

 returning

0.37

 save

0.36

Activations Density 0.014%

programming and database-related terminology, particularly focusing on operations involving data manipulation and persistence

save, update, return

No Comments

No Known Activations