Hostile Neural Networks guide for All the Mods 10: To the Sky 1.6 to 2.0.0
Hostile Neural Networks guide for All the Mods 10: To the Sky
Last verified for All the Mods 10: To the Sky v2.0.0 on Oct 17, 2026 · text last changed Oct 17, 2026
A newer release (2.1.0) is out Guide for the current version
Also applies to:All the Mods 10 0.46 to 4.5
- Sections
- 5
- Recipes shown
- 8
- Items named
- 11
- Quests covered
- 11
4 sections are written for this pack; 1 is shared from another pack's guide.
Every recipe and number here was checked against All the Mods 10: To the Sky 2.0.0's own files.
What is Hostile Neural Networks?
2 parts
Hostile Neural Networks is a mob-simulation mod based on Deep Mob Learning. You train data models on mobs, run simulations with them, and turn the resulting predictions into mob loot. If you want mob drops without hunting the mob every time, this is the mod for you. This pack includes version 6.3.0.[1][2]
How the loop works
- Make a Model Framework and right click the mob you want. It turns into that mob's Data Model.[3]
- Put the model in a Deep Learner and kill that mob. Each Learner holds 4 models. Models start at Faulty and need to reach Basic before you can simulate.[3]
- Run the model in a Simulation Chamber. It needs power and at least one Prediction Matrix. Every simulation gives a generalized prediction, and a successful one also gives a mob prediction.[1][3]
- Feed mob predictions into a Loot Fabricator with power and pick the item you want. It remembers your choice, so you can automate it.[3]
Where it fits in the pack
The mod has its own quest chapter. "Learning Mobs" quest and "Modeling Mobs" quest come first and lead to "Simulating Deaths" quest. "Leveling Up Data Models" quest covers leveling up your models. After Simulating Deaths, the chapter has quests for Overworld, Nether and End generalized predictions, which the quest text says feed various crafting recipes. It also unlocks "Creating Loot" quest, which is followed by a long run of quests that each ask for a mob prediction. The quest examples include Ender Dragons for Dragon Eggs and Withers for Nether Stars.[3]
The mod also appears on the ATM Star path. In the chapter The ATM Star, one quest asks you to hold a Prediction, and it unlocks after Simulating Deaths. So you need a working Simulation Chamber setup to progress there.[4]
Unverified The sources do not say which mob prediction that ATM Star quest needs, or what the ATM Star recipe does with it.
How do you get started with Hostile Neural Networks?
6 parts
Hostile Neural Networks lets you train a model of a mob, simulate it, and turn the result into that mob's drops. The order below follows the pack's own quest chapter, so you can tick the quests off as you go.[1][3]
First, make a Deep Learner
Start with the Deep Learner (quest "Learning Mobs" quest). Right-click it to open a HUD where you store Data Models. Each Learner holds 4 Data Models, and it collects data while it is anywhere in your inventory.[3]
Crafting Shaped
| any obsidian | any obsidian | |
| any glass panes | ||
| any obsidian | any redstone dust | any obsidian |
Recipe id: hostilenetworks:deep_learner
Next, turn a Model Framework into a Data Model
Craft a Model Framework (quest "Modeling Mobs" quest). Right-click the mob you want to study with it, and the framework turns into that mob's Data Model. Put that model in your Deep Learner, then go kill that specific mob.[1][3]
Crafting Shaped
| any redstone dust | any redstone dust | |
| any gold ingot |
Recipe id: hostilenetworks:framework
Level the model up to Basic
A new Data Model starts at the Faulty tier, and you need at least Basic before you can simulate it (quest "Leveling Up Data Models" quest). The Deep Learner HUD and the model's tooltip both show how many kills you still need for the next tier.[1][3]
Build the Simulation Chamber and a Prediction Matrix
The Simulation Chamber is the main machine of the mod (quest "Simulating Deaths" quest). It needs power, and the amount depends on the Data Model inside, which you can read in the model's tooltip. It also needs at least one Prediction Matrix, and the result of the simulation is written onto that matrix.[3]
Crafting Shaped
| Empty slot | any glass panes | Empty slot |
| any obsidian | ||
| any lapi gems | any lapi gems |
Recipe id: hostilenetworks:sim_chamber
Crafting Shaped
| any iron ingot | any glass panes | Empty slot |
| any glass panes | any glass panes | |
| Empty slot | any glass panes | any gold ingot |
Recipe id: hostilenetworks:prediction_matrix
Every simulation gives some kind of Generalized Prediction. If it succeeds, you also get a Mob Prediction for that model. The higher the model's tier, the better the success rate.[1][3]
Cash in with the Loot Fabricator
Finally, the Loot Fabricator (quest "Creating Loot" quest) turns a successful Mob Prediction into actual drops. Give it power and a Mob Prediction, then choose which item you want from that mob's Data Model. It remembers your choice, so you can automate it. The quest gives Zombies for Iron and Withers for Nether Stars as examples.[3]
Crafting Shaped
| Empty slot | any netherite ingot | Empty slot |
| any diamond gem | any obsidian | any diamond gem |
| any gold ingot | any gold ingot |
Recipe id: hostilenetworks:loot_fabricator
Once the basics are running
Generalized Predictions can be used in various crafting recipes. These two in-pack recipes link the Generalized Overworld Prediction, the Generalized Nether Prediction and the Generalized Ender Prediction together.[3][5]
Crafting Shaped
| Empty slot | Empty slot | |
| Empty slot | Empty slot |
Recipe id: hostilenetworks:living_matter/overworld_to_nether
Crafting Shaped
| Empty slot | Empty slot | |
| Empty slot | Empty slot |
Recipe id: hostilenetworks:living_matter/nether_to_ender
Setting up a second Loot Fabricator? A Fabrication Directive stores the selections from a Loot Fabricator. Right-click to copy them, and Shift-Right-click to apply them.[5][6]
Crafting Shaped
| Empty slot | any generalized prediction | Empty slot |
| Empty slot | Empty slot |
Recipe id: hostilenetworks:fab_directive
How does Hostile Neural Networks work?
6 parts
Same as in the All the Mods 10 4.5 guide
Hostile Neural Networks lets you farm mob drops without farming mobs. You train data models, run simulations with them to produce mob predictions, and then turn those predictions into real loot. Every step of that chain has its own item or machine, so let's walk through it in order.[1]
Data Models: the core of the mod
Everything revolves around the Data Model, an item that holds information about one entity and lets you generate that entity's drops. To get one, you first craft a Model Framework and then use it on the mob you want to build a model of.[1]
Crafting Shaped
| any redstone dust | any redstone dust | |
| any gold ingot |
Recipe id: hostilenetworks:framework
The mod's tooltip text includes a line telling you to click on an entity to attune a model, plus a message saying that model attunement is disabled, so watch the tooltip when you try this.[7]
Which mobs have models is data-driven. Custom Data Models can be added through a datapack, with each model defined as a single file in a data_models subfolder.[1]
Training with the Deep Learner
A fresh model starts in the Faulty state and can't be simulated yet. To level it up, put it in a Deep Learner. The Deep Learner holds a few data models at once, and the kills you get start training them. You need to get each model up to Basic this way before the next machine will accept it.[1]
Crafting Shaped
| any obsidian | any obsidian | |
| any glass panes | ||
| any obsidian | any redstone dust | any obsidian |
Recipe id: hostilenetworks:deep_learner
Running simulations in the Simulation Chamber
Once a model is Basic or higher, you can load it into a Simulation Chamber to run simulations. The Simulation Chamber is the second way to train a model, so your models keep improving while they work.[1]
Crafting Shaped
| Empty slot | any glass panes | Empty slot |
| any obsidian | ||
| any lapi gems | any lapi gems |
Recipe id: hostilenetworks:sim_chamber
Each simulation needs energy and a catalyst item. The energy cost depends on the model type. The catalyst is usually a Prediction Matrix, so you'll want a steady supply of those.[1]
Crafting Shaped
| any iron ingot | any glass panes | Empty slot |
| any glass panes | any glass panes | |
| Empty slot | any glass panes | any gold ingot |
Recipe id: hostilenetworks:prediction_matrix
Every simulation produces a generalized prediction. It may also produce a Mob Prediction, depending on the model's accuracy. Higher model tiers mean higher accuracy, up to 99.5% at the Self Aware tier. Leveling your models is what makes a setup truly efficient.[1]
Generalized predictions
The generalized predictions come in three kinds: Generalized Overworld Prediction, Generalized Nether Prediction and Generalized Ender Prediction. In this pack you can craft them up the chain, turning Overworld into Nether and Nether into Ender.[8]
Crafting Shaped
| Empty slot | Empty slot | |
| Empty slot | Empty slot |
Recipe id: hostilenetworks:living_matter/overworld_to_nether
Crafting Shaped
| Empty slot | Empty slot | |
| Empty slot | Empty slot |
Recipe id: hostilenetworks:living_matter/nether_to_ender
Turning predictions into loot
Mob Predictions become real drops in the Loot Fabricator, a powered device. The possible outputs come from a pre-defined list of items for each model type, and you pick a target for each type. The fabricator remembers your choice per type, so you can feed several kinds of predictions into a single fabricator.[1]
Crafting Shaped
| Empty slot | any netherite ingot | Empty slot |
| any diamond gem | any obsidian | any diamond gem |
| any gold ingot | any gold ingot |
Recipe id: hostilenetworks:loot_fabricator
To reuse your picks, grab a Fabrication Directive. It stores selections from a Loot Fabricator: right-click to copy, shift-right-click to apply.[7]
Crafting Shaped
| Empty slot | any generalized prediction | Empty slot |
| Empty slot | Empty slot |
Recipe id: hostilenetworks:fab_directive
What the tooltips tell you
The mod has several info lines that are useful for planning:[7]
- A prompt to hold a key for more info, so check the extended tooltip.[7]
- Generates Experience, followed by a value.[7]
- Loot Table Copies, followed by a value.[7]
- No Base Drop Produced.[7]
Unverified The sources don't give exact energy costs, how much power the Loot Fabricator uses, the kill or simulation counts needed for each tier, the names of the tiers between Basic and Self Aware, or which sides of the machines accept items and energy for automation. Check these in-game before you design an automated setup.
How do you progress through Hostile Neural Networks?
5 parts
Progression in Hostile Neural Networks follows the quest chapter almost step by step: get a Deep Learner and a Model Framework, make a Data Model, level it, then simulate and turn the results into loot. Here's the path in order.[3]
Step 1: Learner and Framework
Two quests open the chapter. "Learning Mobs" quest has you make the Deep Learner. It holds 4 Data Models and collects kill data as long as it's anywhere in your inventory. If you carry several Learners, kills count for the models in all of them.[3]
Crafting Shaped
| any obsidian | any obsidian | |
| any glass panes | ||
| any obsidian | any redstone dust | any obsidian |
Recipe id: hostilenetworks:deep_learner
"Modeling Mobs" quest covers the Model Framework. Right click the mob you want to study with the Framework and it turns into that mob's Data Model. Put that model in a Deep Learner and kill that mob with the Learner in your inventory.[3]
Crafting Shaped
| any redstone dust | any redstone dust | |
| any gold ingot |
Recipe id: hostilenetworks:framework
Step 2: Level up your Data Model
"Leveling Up Data Models" quest covers the Data Model. A new model starts at the first tier, Faulty, and you need to get it to at least Basic before you can simulate with it. The Deep Learner HUD shows how many kills the next level needs, and the model's tooltip shows the same info.[3]
Step 3: Simulate
"Simulating Deaths" quest unlocks the Simulation Chamber, the main machine of the mod. It needs power, and the amount depends on the Data Model inside (check the model's tooltip). It also needs at least one Prediction Matrix, and the simulation result is written onto that Matrix.[3]
Crafting Shaped
| Empty slot | any glass panes | Empty slot |
| any obsidian | ||
| any lapi gems | any lapi gems |
Recipe id: hostilenetworks:sim_chamber
Crafting Shaped
| any iron ingot | any glass panes | Empty slot |
| any glass panes | any glass panes | |
| Empty slot | any glass panes | any gold ingot |
Recipe id: hostilenetworks:prediction_matrix
Every simulation gives some kind of Generalized Prediction, and the type depends on where the mob in the model comes from. If the simulation succeeds, you also get a Mob Prediction for that model. The higher the Data Model's tier, the better the success rate, so leveling pays off.[3]
Generalized Predictions by dimension
Three follow-up quests each hand you one kind of Generalized Prediction. a quest in "Hostile Neural Networks" asks for the Generalized Overworld Prediction, a quest in "Hostile Neural Networks" for the Generalized Nether Prediction, and a quest in "Hostile Neural Networks" for the Generalized Ender Prediction. Their descriptions say they make various Overworld, Nether and End materials.[3]
The pack also lets you convert predictions upward: Overworld to Nether, then Nether to End.[5]
Crafting Shaped
| Empty slot | Empty slot | |
| Empty slot | Empty slot |
Recipe id: hostilenetworks:living_matter/overworld_to_nether
Crafting Shaped
| Empty slot | Empty slot | |
| Empty slot | Empty slot |
Recipe id: hostilenetworks:living_matter/nether_to_ender
Step 4: Loot
"Creating Loot" quest gives you the Loot Fabricator. Give it power and a successful Mob Prediction and you choose which item you want from that mob's Data Model. It remembers your choice, so you can automate it. The quest's examples are Ender Dragons for Dragon Eggs, Zombies for Iron, and Withers for Nether Stars.[3]
Crafting Shaped
| Empty slot | any netherite ingot | Empty slot |
| any diamond gem | any obsidian | any diamond gem |
| any gold ingot | any gold ingot |
Recipe id: hostilenetworks:loot_fabricator
After this, the chapter continues with a long run of quests that each ask for a Prediction, one per mob you model. A separate quest in the ATM Star chapter, a quest in "Chapter 4: The ATM Star", also asks for a Mob Prediction and unlocks after Simulating Deaths, so the simulation setup counts toward your main progression too.[3][4]
Unverified The quests only name Faulty and Basic, so the names of the higher Data Model tiers and the kill counts for each level aren't confirmed here. Check the Deep Learner HUD or the model tooltip in game for the exact numbers.
Which All the Mods 10: To the Sky quests use Hostile Neural Networks?
3 parts
Most of this mod's quests sit in a single chapter called Hostile Neural Networks. One more quest that asks for a prediction item shows up in Chapter 4: The ATM Star. The chapter starts with an intro quest, Welcome to Hostile Neural Networks!, and everything else branches from there.[3][4]
Getting started
- "Learning Mobs" quest asks you to get a Deep Learner. It opens straight after the intro quest. The Deep Learner holds 4 Data Models and collects kill data as long as it is anywhere in your inventory.[3]
- "Modeling Mobs" quest also opens after the intro quest and asks for a Model Framework. Right click a mob with it to turn it into that mob's Data Model, then put the model in a Deep Learner and kill that mob.[3]
- "Leveling Up Data Models" quest asks for a Data Model and unlocks once you've done both quests above. Data Models start at the Faulty tier, and the quest text says you need to reach at least Basic before you can run simulations.[3]
- "Simulating Deaths" quest asks for a Simulation Chamber and a Prediction Matrix. It also unlocks after Learning Mobs and Modeling Mobs. The quest text says the chamber needs power and at least one Prediction Matrix to run a simulation, and that a higher Data Model tier gives a better success rate.[3]
After Simulating Deaths
Finishing Simulating Deaths opens several quests at once. Four of them are about predictions and loot:[3]
- a quest in "Hostile Neural Networks" asks for a Generalized Overworld Prediction. Its description says it makes various Overworld materials.[3]
- a quest in "Hostile Neural Networks" asks for a Generalized Nether Prediction. Its description says it makes various Nether materials.[3]
- a quest in "Hostile Neural Networks" asks for a Generalized Ender Prediction. Its description says it makes various End materials.[3]
- "Creating Loot" quest asks for a Loot Fabricator. The quest text says it turns a successful Mob Prediction into loot, remembers the item you pick so it can be automated, and gives examples such as Zombies for Iron, Withers for Nether Stars and Ender Dragons for Dragon Eggs.[3]
The prediction quests
Creating Loot leads into a long run of untitled quests. Each one asks you to hand in a Prediction, and they all unlock after Creating Loot. Examples are a quest in "Hostile Neural Networks" and a quest in "Hostile Neural Networks". The sources don't say which mob each one is for.[3]
Chapter 4: The ATM Star has one related quest, a quest in "Chapter 4: The ATM Star". It also asks for a Prediction and unlocks after Simulating Deaths.[4]
Sources
- Hostile Neural Networks project page (modrinth) · official mod data
- All the Mods 10: To the Sky 2.0.0 mod list · pack file
- All the Mods 10: To the Sky 2.0.0 quests: "Hostile Neural Networks" · pack file
- All the Mods 10: To the Sky ATM Star path quests in "Chapter 4: The ATM Star" · pack file
- All the Mods 10: To the Sky 2.0.0 recipe dump: recipes making Hostile Neural Networks items · pack recipe dump
- hostilenetworks tooltips and descriptions (English lang file) · official mod data
- hostilenetworks tooltips and descriptions (English lang file) · official mod data
- All the Mods 10 4.5 recipe dump: recipes making Hostile Neural Networks items · pack recipe dump
Where this guide comes from
Written from what All the Mods 10: To the Sky itself ships: its files and recipes, the in-game guidebooks and the mods' own pages. Every statement links to the source it came from, and the numbers are checked against the pack before publishing.
Guide last verified Oct 17, 2026.
Guides for other versions: 2.0.3 to 2.1.0 · 2.0.1 to 2.0.2 · 1.1.1 to 1.2.1 · 1.0 to 1.0.2