
Hostile Neural Networks in All the Mods 11
All the Mods 11 0.12.0-beta ships version 7.0.0
Hostile Neural Networks is a mod in All the Mods 11 0.12.0-beta, where it is version 7.0.0. Simulate Mobs to create not-so-simulated drops! (Summary from the mod's Modrinth page.)
Changed in 0.12.0-beta
10 new recipes · 14 new items
Compared with 0.10.0-beta, the last version we have recipe data for. Everything that changed in 0.12.0-beta →
Hostile Neural Networks guide for All the Mods 11
Last verified for All the Mods 11 v0.12.0-beta on Oct 17, 2026 · text last changed Oct 15, 2026
Also applies to:FTB Evolution 2 1.0.2 to 1.4.0
- Sections
- 5
- Recipes shown
- 10
- Items named
- 9
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 11 0.12.0-beta, read from the running game on Oct 15, 2026.
What is Hostile Neural Networks?
2 parts
Same as in the FTB Evolution 2 1.1.0 guide
Hostile Neural Networks lets you simulate mobs to get their drops without hunting them. It is based on Deep Mob Learning. You train data models, run simulations with them, and turn the resulting predictions into mob loot. This pack ships version 7.0.0 of the mod.[1][2]
How the loop works
The core idea is the Data Model, an item that holds information about one type of entity. To get one, you craft a Data Model Framework and use it on the target entity to build the model. After that, you have to train it.[1]
- Build a Data Model from a Data Model Framework by using the framework on the target entity.[1]
- Train the model. You can do this with a Deep Learner or with a Simulation Chamber. The Deep Learner holds a few models, and the kills you make begin to train them. Every model starts out Faulty and has to be trained up to Basic in the Deep Learner.[1]
- Once the model is Basic or higher, run simulations in the Simulation Chamber. This needs energy (based on the model type) and a catalyst item, which is usually a Prediction Matrix.[1]
- Feed the Mob Predictions into a Loot Fabricator. It uses power to turn predictions into mob drops.[1]
Simulations always give a generalized prediction, and they may also give a Mob Prediction depending on how accurate the model is. Higher model tiers are more accurate, up to 99.5% at Self Aware.[1]
Custom models
Data Models can also be added with a datapack. Each model is a single file in the data_models subfolder.[1]
How do you get started with Hostile Neural Networks?
5 parts
Hostile Neural Networks lets you simulate mobs and turn the results into real mob drops. The whole mod is built around the Data Model, an item that holds information about one type of entity. Here is the order I'd go in, and why.[1]
Step one: make a Model Framework
Everything starts with the Model Framework. You use it on the mob you want to target to build that mob's Data Model.[1]
Crafting Shaped
| Clay Ball | Repeater | Clay Ball |
| any redstone dust | Smooth Stone | any redstone dust |
| Clay Ball | any gold ingot | Clay Ball |
Recipe id: hostilenetworks:framework
Use the finished framework on the target entity to build its Data Model.
Step two: craft a Deep Learner
A fresh model is no use yet. Every model starts in the Faulty state and has to be trained up to Basic in the Deep Learner. It holds a few data models, and the kills you get start training them.[1]
Crafting Shaped
| any obsidian | Repeater | any obsidian |
| Repeater | any glass panes | Repeater |
| any obsidian | any redstone dust | any obsidian |
Recipe id: hostilenetworks:deep_learner
Step three: build the Simulation Chamber and Prediction Matrices
Once a model is Basic or higher, you can run simulations with the Simulation Chamber. Simulations always produce a generalized prediction, and they may also produce a Mob Prediction. The chance depends on the model's accuracy, which rises with its tier, up to 99.5% at Self Aware.[1][4]
Crafting Shaped
| Empty slot | any glass panes | Empty slot |
| Ender Pearl | any obsidian | Ender Pearl |
| any lapi gems | Comparator | any lapi gems |
Recipe id: hostilenetworks:sim_chamber
Simulations cost energy, which depends on the model type, and they need a catalyst item. That is usually the Prediction Matrix, so craft some before you start.[1]
Crafting Shaped
| any iron ingot | any glass panes | Empty slot |
| any glass panes | Clay Ball | any glass panes |
| Empty slot | any glass panes | any gold ingot |
Recipe id: hostilenetworks:prediction_matrix
One craft gives you 16 Prediction Matrices.
Step four: set up a Loot Fabricator
Mob Predictions are not the loot itself. You process them in the Loot Fabricator, which uses power to materialize predictions into real items. You choose a target drop for each type of model, and the fabricator remembers your choice for each type. That means one fabricator can take several kinds of prediction.[1][4]
Crafting Shaped
| Empty slot | any netherite ingot | Empty slot |
| any diamond gem | any obsidian | any diamond gem |
| any gold ingot | Comparator | any gold ingot |
Recipe id: hostilenetworks:loot_fabricator
The Fabrication Directive stores selections from a Loot Fabricator. Right-click to copy, and Shift-Right-Click to apply. It's handy when you want to reuse a layout.[4]
Crafting Shaped
| Empty slot | any generalized prediction | Empty slot |
| Paper | Paper | |
| Empty slot | Paper | Empty slot |
Recipe id: hostilenetworks:fab_directive
Later: the Data Center
When you want automation, the Data Center runs up to 25 Data Models at once. Each active slot consumes one input per cycle and produces the drop chosen by its Fabrication Directive.[4]
Crafting Shaped
| Echo Shard | Nether Star | Echo Shard |
| any obsidian | ||
| any obsidian | any obsidian | any obsidian |
Recipe id: hostilenetworks:data_center
It needs a hollow 7×7×7 shell. The floor must be obsidian, and the walls and ceiling must be black stained glass. The controller replaces one bottom-wall block and must face outwards.[4]
The Data Center IO Port can replace any wall or ceiling block in the shell. Right-click it to cycle its mode between Energy, Models, Inputs and Outputs.[4]
Crafting Shaped
| any iron ingot | Black Stained Glass | any iron ingot |
| Black Stained Glass | Redstone Block | Black Stained Glass |
| any iron ingot | Black Stained Glass | any iron ingot |
Recipe id: hostilenetworks:data_center_io_port
How does Hostile Neural Networks work?
6 parts
Hostile Neural Networks, based on Deep Mob Learning, turns mobs into loot through a simple chain. You build a Data Model of a mob, train it, run simulations to get predictions, then turn those predictions into real drops. Here is how each machine in that chain works in this pack.[1]
Building and training a Data Model
The Data Model is the core unit of the mod. You craft a Model Framework, then use it on the target entity to build the model.[1]
Crafting Shaped
| Clay Ball | Repeater | Clay Ball |
| any redstone dust | Smooth Stone | any redstone dust |
| Clay Ball | any gold ingot | Clay Ball |
Recipe id: hostilenetworks:framework
The Model Framework is the blank you use on a mob.
Every model starts in the Faulty state. The Deep Learner holds a few models, and kills you make begin to train them. You must train a model up to Basic in the Deep Learner before it can be used in a simulation. The pack also adds a keybind category for the mod with an option to open the Deep Learner as a Curio.[1][4]
Crafting Shaped
| any obsidian | Repeater | any obsidian |
| Repeater | any glass panes | Repeater |
| any obsidian | any redstone dust | any obsidian |
Recipe id: hostilenetworks:deep_learner
The Simulation Chamber
Once a model is Basic or higher, the Simulation Chamber can run simulations of it. A simulation always produces a generalized prediction, and it may also produce a Mob Prediction. The chance depends on the model's accuracy, which rises with the model tier, up to 99.5% at Self Aware.[1]
A simulation needs energy, based on the model type, plus a catalyst item. The catalyst is usually a Prediction Matrix. Every Eternal Starlight model this pack adds uses the Prediction Matrix as its input.[1][5]
Crafting Shaped
| Empty slot | any glass panes | Empty slot |
| Ender Pearl | any obsidian | Ender Pearl |
| any lapi gems | Comparator | any lapi gems |
Recipe id: hostilenetworks:sim_chamber
Crafting Shaped
| any iron ingot | any glass panes | Empty slot |
| any glass panes | Clay Ball | any glass panes |
| Empty slot | any glass panes | any gold ingot |
Recipe id: hostilenetworks:prediction_matrix
One craft makes 16 Prediction Matrix.
The chamber has two modes. In training mode, simulations increase model data but produce no outputs. In inference mode, simulations produce output but the models gain no data. Pick the mode that matches what you want from the run.[4]
The Loot Fabricator
Mob Predictions are processed in the Loot Fabricator, which uses power to convert predictions into mob drops. The available drops come from a predefined list, and you pick a target for each type of model. The fabricator remembers your choice per type, so one fabricator can take several types of prediction.[1][4]
Crafting Shaped
| Empty slot | any netherite ingot | Empty slot |
| any diamond gem | any obsidian | any diamond gem |
| any gold ingot | Comparator | any gold ingot |
Recipe id: hostilenetworks:loot_fabricator
If you set up several fabricators, a Fabrication Directive saves you the clicking. Right-click a fabricator with it to copy its selections, then Shift-Right-click another to apply them.[4]
Crafting Shaped
| Empty slot | any generalized prediction | Empty slot |
| Paper | Paper | |
| Empty slot | Paper | Empty slot |
Recipe id: hostilenetworks:fab_directive
The mod also has Overworld, Nether, Twilight and Ender generalized predictions. The pack has crafting recipes for the Nether and Ender ones.[3][4]
Crafting Shaped
| Empty slot | Overworld Prediction | Empty slot |
| Overworld Prediction | Netherrack | Overworld Prediction |
| Empty slot | Overworld Prediction | Empty slot |
Recipe id: hostilenetworks:living_matter/overworld_to_nether
Crafting Shaped
| Empty slot | Empty slot | |
| End Stone | ||
| Empty slot | Empty slot |
Recipe id: hostilenetworks:living_matter/nether_to_ender
The Data Center multiblock
The Data Center is the automation endgame of the mod. It runs up to 25 Data Models at once, each producing one drop per cycle into a shared output buffer.[4]
To operate it, put Data Models in the 5×5 grid and matching items in the four input slots. Each active slot uses one input per cycle and produces the drop chosen by its Fabrication Directive.[4]
It needs a 7×7×7 hollow shell. Here is what the shell is made of:[4]
- The floor must be obsidian. The floor block tag accepts the common obsidians tag.[4][6]
- The walls and ceiling must be black stained glass.[4]
- The controller replaces one bottom-wall block and must face outwards.[4]
Crafting Shaped
| Echo Shard | Nether Star | Echo Shard |
| any obsidian | ||
| any obsidian | any obsidian | any obsidian |
Recipe id: hostilenetworks:data_center
Data Center IO Ports
The Data Center IO Port is how you automate the Data Center. It can replace any wall or ceiling block in the shell, and the controller detects it automatically.[4]
Right-click a port to cycle its mode between Energy, Models, Inputs and Outputs. Energy accepts power, while Models, Inputs and Outputs expose the matching item slots. Pipes, hoppers or cables can then feed or empty the Data Center through the port.[4]
Crafting Shaped
| any iron ingot | Black Stained Glass | any iron ingot |
| Black Stained Glass | Redstone Block | Black Stained Glass |
| any iron ingot | Black Stained Glass | any iron ingot |
Recipe id: hostilenetworks:data_center_io_port
What this pack changes
The pack uses a KubeJS script to add Data Models for Eternal Starlight mobs, and it only does so when that mod is loaded. Each model has a sim_cost of 128, 256, 512 or 8,192, uses the Prediction Matrix as its input, and has a base drop of a pack-made starlight prediction item. Each model also lists its own set of fabricator drops with fixed counts.[5]
Two of the models, Lunar Monstrosity and Starlight Golem, have the highest sim_cost at 8,192 and the longest fabricator drop lists. Both also define data per kill: 3 for faulty, 12 for basic, 30 for advanced and 45 for superior.[5]
How do you progress through Hostile Neural Networks?
4 parts
No quests in this pack involve Hostile Neural Networks, and the mod is not on the ATM Star quest path. So there is no quest chain to follow here. Your progression comes from the machines and from the tiers of Generalized Predictions, using the recipes the pack ships.[3]
Starter items and machines
The Prediction Matrix recipe makes 16 per craft, so one craft stocks you up for a while.[3]
Crafting Shaped
| Clay Ball | Repeater | Clay Ball |
| any redstone dust | Smooth Stone | any redstone dust |
| Clay Ball | any gold ingot | Clay Ball |
Recipe id: hostilenetworks:framework
The Model Framework.
Crafting Shaped
| any iron ingot | any glass panes | Empty slot |
| any glass panes | Clay Ball | any glass panes |
| Empty slot | any glass panes | any gold ingot |
Recipe id: hostilenetworks:prediction_matrix
One craft gives you 16 Prediction Matrix.
Crafting Shaped
| any obsidian | Repeater | any obsidian |
| Repeater | any glass panes | Repeater |
| any obsidian | any redstone dust | any obsidian |
Recipe id: hostilenetworks:deep_learner
The Deep Learner.
Crafting Shaped
| Empty slot | any glass panes | Empty slot |
| Ender Pearl | any obsidian | Ender Pearl |
| any lapi gems | Comparator | any lapi gems |
Recipe id: hostilenetworks:sim_chamber
The Simulation Chamber.
Crafting Shaped
| Empty slot | any netherite ingot | Empty slot |
| any diamond gem | any obsidian | any diamond gem |
| any gold ingot | Comparator | any gold ingot |
Recipe id: hostilenetworks:loot_fabricator
The Loot Fabricator.
Crafting Shaped
| Empty slot | any generalized prediction | Empty slot |
| Paper | Paper | |
| Empty slot | Paper | Empty slot |
Recipe id: hostilenetworks:fab_directive
The Fabrication Directive.
Building the Data Center
The Data Center has block tags for its floor and walls. The floor accepts blocks from the obsidians tag. The walls accept black stained glass or the Data Center IO Port, and several black glass variants from other mods are accepted as optional wall blocks.[6][7]
Crafting Shaped
| Echo Shard | Nether Star | Echo Shard |
| any obsidian | ||
| any obsidian | any obsidian | any obsidian |
Recipe id: hostilenetworks:data_center
The Data Center.
Crafting Shaped
| any iron ingot | Black Stained Glass | any iron ingot |
| Black Stained Glass | Redstone Block | Black Stained Glass |
| any iron ingot | Black Stained Glass | any iron ingot |
Recipe id: hostilenetworks:data_center_io_port
The IO Port counts as a wall block.
Prediction tiers
The mod groups its Generalized Predictions into Overworld, Nether, End and Twilight tiers. This pack has crafting recipes for the Nether and Ender ones, filed under living_matter.[3][8]
Crafting Shaped
| Empty slot | Overworld Prediction | Empty slot |
| Overworld Prediction | Netherrack | Overworld Prediction |
| Empty slot | Overworld Prediction | Empty slot |
Recipe id: hostilenetworks:living_matter/overworld_to_nether
Crafts the Generalized Nether Prediction.
Crafting Shaped
| Empty slot | Empty slot | |
| End Stone | ||
| Empty slot | Empty slot |
Recipe id: hostilenetworks:living_matter/nether_to_ender
Crafts the Generalized Ender Prediction.
What this pack changes
The pack adds its own shaped recipe for the Dragon Head. It is built around the Generalized Ender Prediction and a prediction for the Wither data model, so the End tier has a use beyond the machines.[9]
When Enderman Overhaul is loaded, the pack also generates its own Enderman data model, covering the base Enderman and the Enderman Overhaul variants. Its simulation cost is 512, its input is the Prediction Matrix, and its base drop is the Generalized Ender Prediction.[9]
- The Loot Fabricator drops for the Enderman model include 16 Ender Pearls and 1 End Crystal.[9]
- Some further drops are marked optional: 4 Ender Tears from EvilCraft, 2 Nebulous Hearts from Reliquary Reincarnations and 4 Enderman Heads from Ender IO.[9]
What changed in Hostile Neural Networks between versions?
All the Mods 11 0.12.0-beta ships Hostile Neural Networks version 7.0.0, using the file HostileNeuralNetworks-26.1.2-7.0.0.jar. The pack's mod list says it was first included in 0.12.0-beta or an earlier release, so this is the version you are playing with.[10]
Unverified What changed in Hostile Neural Networks between releases is not shown in the sources here. The pack changelogs have no entries for this mod, and the mod's 7.0.0 changelog text is not included, so no notable changes are listed in this section.
Sources
- Hostile Neural Networks project page (modrinth) · official mod data
- FTB Evolution 2 1.1.0 mod list · pack file
- All the Mods 11 0.12.0-beta recipe dump: recipes making Hostile Neural Networks items · pack recipe dump
- hostilenetworks names, screen text and descriptions (English lang file) · official mod data
- kubejs/server_scripts/mods/eternal_starlight/datapacks.js (All the Mods 11's own script) · pack file
- data/hostilenetworks/tags/block/data_center_floor.json (hostilenetworks's jar) · official mod data
- data/hostilenetworks/tags/block/data_center_wall.json (hostilenetworks's jar) · official mod data
- data/hostilenetworks/tags/item/generalized_predictions.json (hostilenetworks's jar) · official mod data
- kubejs/server_scripts/mods/hostilenetworks/recipes.js (All the Mods 11's own script) · pack file
- All the Mods 11 0.12.0-beta mod list · pack file
Where this guide comes from
Written from what All the Mods 11 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.
Report wrong info
Version in each release
1 release, first in 0.12.0-beta
| All the Mods 11 | Version | Jar | Where it came from |
|---|---|---|---|
| 0.12.0-beta Oct 15, 2026 | 7.0.0 | HostileNeuralNetworks-26.1.2-7.0.0.jar | From the published release |
"From pack files" means the list came from the pack's own files; "from the published release" means ForgeBase read the jars of the release itself.
Comments and tips
Loading comments...