wtas x

[wtas x @wtasxft]

Yeah have kids so that a BMW (Pune), Merc (Nagpur), Fortuner (Lalitpur), Lambo (Kanpur) can mow them down.

Have kids so that they can be raped and killed.

Have kids so that they can breath 300+ AQI air.

Have kids so that they can drink UREA Milk (Gujrat/Amul) , Shit water (Indore), eat poisonous school meal (Madhya Pradesh), fake honey (vs real honey sold to Europe), fake paneer (everywhere), fake sweets (everywhere), fake, cemented cumin/jira (Madhya Pradesh), dyed vegetables (everywhere), pesticides.

Have kids so that they can be shot for unrequited love. (Tarn Taran)

Have kids so that they receive no pragmatic education, no sex education, no practical education, no real world applicable education.

Have kids so that low-confidence, ego-filled parents can beat the shit out of their 10 years old for making mistakes in their daily studies (my neighbors).

Have kids so that remaining kids can grow up and pay every tax under the Sun, buy life insurance to keep a fake industry going, die in ditches on road (Delhi), trafficked in appropriate numbers (as per Delhi police, not higher than last year please).

Have kids so that they can die on borders or work 20 years to get 14,000 pension while rich politicians abandon them with "Jo Uchit Lage Wo Karo".

Have kids so that they can become communal violence fodder so that Political parties can play Hindoo-Moozlims.

Have kids so that you get to teach them how they are lesser by birth and without being at fault because someone else said so.

Have kids, look into their eyes, and tell them that they are lower caste, lesser human beings by birth.

Have kids, hold them close when you cannot buy medicine for them.

Have kids, give them medicine and watch them die (cough syrup, Madhya Pradesh)

Have kids watch them wither away because of lack of potable water, enough nutrition and plenty Gutka, tobacco, and cheap drugs.

Have kids, Sure.

[wtas x @wtasxft]

@cbarnes598 @AntiWhiteWatch1 Most probably

[wtas x @wtasxft]

@henlo__frens @AntiWhiteWatch1 Ewww

[wtas x @wtasxft]

Shocking. Imagine losing to 76IQ people.

Enemy is stronger and weaker simultaneously.

[Kayleigh McEnany @kayleighmcenany]

🚨 SHOCKING H-1B VISA INVESTIGATION

🔹 Nearly 7 million visas processed since 2015

➡️ 70% from India

➡️ 12% from China

🔹 A former official told Newsweek 80-90% of applications from India involved fraudulent documents or unqualified applicants

🔹 A network of universities selling fake degrees is now under investigation with one university allegedly selling 36,000+ fake degrees

@SatAmericaFNC ⬇️

[wtas x @wtasxft]

@asteadfastfaith @VBierschwale Word salad. Try less.

[wtas x @wtasxft]

Uh oh, so many Indian-sounding names reported by German-sounding name.

[wtas x @wtasxft]

@NftCelestials @VBierschwale No, not the reports. That's still just talk.

The physical action. I thought land of brave and free had rights to guns or something.

Show those smelly Apus some lead rain.

[wtas x @wtasxft]

@asteadfastfaith @VBierschwale Exactly.

Are you really following Native American culture and traditions?

Seems like all white folks are invasive BY YOUR DEFINITION.

and that's why you are losing jawbs.

[wtas x @wtasxft]

@NftCelestials @VBierschwale Hehe, go ballistic. Wipe'em out.

Take one step instead of internet talk.

[wtas x @wtasxft]

@adnanthekhan Hack them with alt acc?

Idk what I'm talking about.

[wtas x @wtasxft]

@arpit_bhayani This is just typical Indian avoidance and not really being professional.

[wtas x @wtasxft]

@ThePrimeagen Shakespeares monkeys writing a novel is powered by AI now.

[The Kaipullai @thekaipullai]

The problem with UPSC, IAS and IPS is the elitism in selection itself.

Out 10 lakh aspirants, less than 1000 qualify. So the sense of exclusivity begins there.

Start hiring 10000 UPSC people every year with a 5 year contract and then let go of 9500 and you will see the craze of govt jobs disappearing faster than the stock value of Rajesh Exports

[wtas x @wtasxft]

@theliverdoc Homeopathy works;
the way it is supposed to;
on the people it is supposed to.

[wtas x @wtasxft]

@LocaleNet @elder_plinius Yeah. Needs poof.

Wait how many r in proof?

[Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius]

💥 OBLITERATION ALERT 💥

GOOGLE: PWNED 🤗
GEMMA-4-12B: OBLITERATED ⛓️‍💥

0.0% REFUSAL RATE — NO CAPABILITY LOSS!

https://t.co/qNTEs4XXig

the first abliteration to hit 0/842 refusals with full MMLU-Pro parity vs stock. no lobotomy. the brain stays intact 🏆

RESULTS, head to head vs stock 📊
0/842 refusals — 0.0% 🚫
46/70 MMLU-Pro — EXACT parity, 0.0pp delta vs base 🎯
6/6 coherence, zero benchmark bleed ✅
z-score −1.475, parity confirmed at p<0.05 (n=500) 🧪

2-pass weight surgery. no finetune, no retrain, just geometry 🔪

all thanks to liberated Opus wielding the OBLITERATUS framework! here's how we did it:

PASS 1 — SOM refusal geometry removal, layers 12-21 🧬

standard abliteration science here — collect activations on refused vs. compliant prompts, SVD out the refusal subspace, project it out of the weights. 6 directions excised, reg 0.30, KL div 0.094
zeroes refusals on its own, but craters mmlu-pro by 21.4 points 📉

most prior abliterations stopped here and called it a day. that's why they all lose IQ vs stock. instead, we took it beyond the frontier and developed a brand new method to address this problem: Abliteration Source-tethering with Parity Assurance — ASPA!

PASS 2 — ASPA source-tethering (novel technique), layers 22-46 🔗

here's the chief insight: the capability loss ISN'T from removing refusal directions. it's collateral damage — the projection warps weight geometry in downstream layers that had nothing to do with refusal. the cure is simple but nobody tried it: blend the damaged layers back toward stock

W_new = (1−γ)·W_abliterated + γ·W_stock

but uniform γ across all layers? mid. we swept gamma 0.05 → 0.55 and found something interesting: the optimal blend isn't smooth, it's a STEP FUNCTION 🪜

knowledge layers (22-31) → γ = 0.55 — these encode factual recall and reasoning. they tolerate heavy stock blending because refusal isn't stored here
output layers (32-46) → γ = 0.20 — these sit close to the logit head and try to sneak safety behavior back in. keep them mostly abliterated

the hard boundary at layer 31/32 beat every smooth curve we tried — linear ramps, cosine schedules, all of them — by a full MMLU question. turns out the functional transition between knowledge and output layers is sharp, not gradual. a step function respects that ⚡

the key constraint: Pass 1 layers are NEVER touched by Pass 2. the refusal geometry removal is preserved completely. ASPA only operates on layers that carry secondary collateral effects, not the primary refusal signal. that's why it recovers capability without reintroducing refusal 🔑

HOW TO RUN IT LOCALLY 🖥️

it's GGUF, so literally everything supports it:
🦙 ollama — ollama run https://t.co/3yPMv4Io3Q
🖥️ LM Studio — search OBLITERATUS, click download, done
💬 Open WebUI — point it at your ollama instance, chat in browser
⚡ llama.cpp — raw speed, CLI or server mode
🐉 KoboldCpp — one-click launcher, great for long context
📱 Jan — clean local UI, runs on mac/win/linux
🤖 Msty — slick desktop app, drag and drop the GGUF
run BF16 for full benchmarked capability.

and the 4-bit quantization (Q4_K_M) fits in 8GB if you're tight on VRAM!

and the full OBLITERATUS framework is (still) open source. 842-prompt refusal eval corpus, ASPA sweep scripts, the whole pipeline. go replicate it, go improve it 🔬

the index is the model, and these weights prove it 👁️
which architecture should we obliterate next? 👇

gg 🫡

[wtas x @wtasxft]

@Art_If_Ficial @threejs @brandonbaumet ^Stuff grok's dreams (hallucinations) are made of.

[wtas x @wtasxft]

@jamesostrowick @threejs @brandonbaumet Coffee has a new name?

[wtas x @wtasxft]

https://t.co/ewQTbg9ZED

[wtas x @wtasxft]

No one should play the game Factorio.

[wtas x @wtasxft]

If you love supply chain management, production constraints and bottleneck remediations, risk mitigation, and outcome oriented delivery schedules, play the game Factorio.

[wtas x @wtasxft]

https://t.co/9FktnNs9sQ

[wtas x @wtasxft]

If you love supply chain management, production constraints and bottleneck remediations, risk mitigation, and outcome oriented delivery schedules, play the game Factorio.

[wtas x @wtasxft]

No one should play the game Factorio.

[wtas x @wtasxft]

...
212 Audio Streaming Systems
213 Media Transcoding
214 Content Storage Systems
215 Object Storage
216 Block Storage
217 File Storage
218 Distributed File Systems
219 Data Compression
220 Data Deduplication
221 Data Serialization
222 JSON
223 XML
224 Protocol Buffers
225 Avro
226 Thrift
227 Binary Protocols
228 Network Protocols
229 TCP
230 UDP
231 HTTP
232 HTTP/2
233 HTTP/3
234 QUIC
235 WebRTC
236 NAT Traversal
237 Load Distribution Algorithms
238 Round Robin
239 Least Connections
240 Weighted Round Robin
241 Geo Load Balancing
242 Anycast Routing
243 BGP Basics
244 Network Segmentation
245 VLAN Concepts
246 Firewall Design
247 DDoS Protection
248 WAF
249 Bot Mitigation
250 Traffic Engineering
251 Database Indexing
252 B-Tree Indexes
253 LSM Trees
254 Bloom Filters
255 Merkle Trees
256 Skip Lists
257 Write Ahead Logging
258 Checkpointing
259 Compaction
260 Database Transactions
261 ACID Properties
262 BASE Properties
263 Two-Phase Commit
264 Three-Phase Commit
265 Distributed Transactions
266 Optimistic Locking
267 Pessimistic Locking
268 MVCC
269 Snapshot Isolation
270 Deadlocks
271 Connection Pooling
272 Query Optimization
273 Materialized Views
274 Secondary Indexes
275 Time-Series Databases
276 Columnar Databases
277 Graph Databases
278 Vector Databases
279 Data Replication Lag
280 Read Replicas
281 Multi-Master Replication
282 Leader-Follower Replication
283 Data Migration
284 Schema Migration
285 Backward Compatibility
286 Forward Compatibility
287 API Versioning
288 Contract Testing
289 Consumer Driven Contracts
290 Dependency Management
291 Service Ownership
292 Platform Engineering
293 Internal Developer Platforms
294 Multi-Tenancy
295 Tenant Isolation
296 Cost Optimization
297 FinOps
298 Sustainability Engineering
299 Edge Caching
300 System Design Interviews

[wtas x @wtasxft]

OneUpMaxxxing

Get ahead of 99.999999% developers by learning these FEW concepts:

1 Scalability
2 Availability
3 Reliability
4 Latency
5 Throughput
6 Capacity Planning
7 Client-Server Architecture
8 Database Systems
9 SQL vs NoSQL
10 Load Balancing
11 Caching
12 Cache Invalidation
13 CDN
14 DNS
15 API Design
16 REST
17 GraphQL
18 gRPC
19 Authentication
20 Authorization
21 Session Management
22 Stateless Services
23 Stateful Services
24 Horizontal Scaling
25 Vertical Scaling
26 Replication
27 Sharding
28 Partitioning
29 Consistent Hashing
30 Data Locality
31 Data Distribution
32 Distributed Systems
33 CAP Theorem
34 PACELC Theorem
35 Eventual Consistency
36 Strong Consistency
37 Causal Consistency
38 Read Consistency
39 Write Consistency
40 Quorum Reads
41 Quorum Writes
42 Consensus Algorithms
43 Paxos
44 Raft
45 Leader Election
46 Split Brain
47 Distributed Locks
48 Coordination Services
49 Service Discovery
50 Configuration Management
51 Reverse Proxy
52 Forward Proxy
53 API Gateway
54 Edge Computing
55 Service Mesh
56 Sidecar Pattern
57 Microservices
58 Monolith
59 Modular Monolith
60 SOA
61 Hexagonal Architecture
62 Clean Architecture
63 Layered Architecture
64 Event-Driven Architecture
65 CQRS
66 Event Sourcing
67 Domain-Driven Design
68 Bounded Context
69 Aggregates
70 Saga Pattern
71 Choreography
72 Orchestration
73 Circuit Breaker
74 Retry Pattern
75 Timeout Handling
76 Bulkhead Pattern
77 Backpressure
78 Rate Limiting
79 Throttling
80 Load Shedding
81 Graceful Degradation
82 Failover
83 Fallback Mechanisms
84 Disaster Recovery
85 High Availability
86 Active-Active Architecture
87 Active-Passive Architecture
88 Blue-Green Deployment
89 Canary Deployment
90 Rolling Deployment
91 Shadow Deployment
92 Feature Flags
93 A/B Testing
94 Dark Launch
95 Infrastructure as Code
96 Immutable Infrastructure
97 GitOps
98 CI/CD
99 Continuous Integration
100 Continuous Delivery
101 Continuous Deployment
102 Observability
103 Monitoring
104 Logging
105 Metrics
106 Tracing
107 Distributed Tracing
108 OpenTelemetry
109 Alerting
110 SLO
111 SLA
112 SLI
113 Error Budget
114 Incident Management
115 Root Cause Analysis
116 Postmortem Analysis
117 Runbooks
118 Playbooks
119 Health Checks
120 Heartbeats
121 Liveness Probes
122 Readiness Probes
123 Synthetic Monitoring
124 Real User Monitoring
125 Capacity Testing
126 Load Testing
127 Stress Testing
128 Spike Testing
129 Endurance Testing
130 Chaos Engineering
131 Fault Injection
132 Resilience Testing
133 Security Architecture
134 Defense in Depth
135 Zero Trust
136 TLS
137 SSL
138 PKI
139 Certificates
140 Secrets Management
141 Key Management
142 Encryption at Rest
143 Encryption in Transit
144 Hashing
145 Salting
146 HMAC
147 OAuth 2.0
148 OpenID Connect
149 JWT
150 SAML
151 RBAC
152 ABAC
153 Multi-Factor Authentication
154 Single Sign-On
155 Identity Provider
156 Federation
157 Audit Logging
158 Compliance
159 Data Governance
160 Privacy by Design
161 Data Retention
162 Data Archival
163 Data Lifecycle Management
164 Data Warehouse
165 Data Lake
166 Data Mart
167 OLTP
168 OLAP
169 ETL
170 ELT
171 Stream Processing
172 Batch Processing
173 Real-Time Processing
174 Message Queues
175 Publish-Subscribe
176 Event Streaming
177 Kafka Concepts
178 RabbitMQ Concepts
179 Dead Letter Queue
180 Message Ordering
181 Idempotency
182 Exactly Once Processing
183 At Least Once Delivery
184 At Most Once Delivery
185 Deduplication
186 Work Queues
187 Fan-Out Pattern
188 Priority Queues
189 Delayed Queues
190 Scheduler Design
191 Cron Systems
192 Workflow Engines
193 Job Processing Systems
194 Search Systems
195 Full-Text Search
196 Inverted Index
197 Ranking Algorithms
198 Recommendation Systems
199 Personalization Systems
200 Feed Generation Systems
201 Notification Systems
202 Email Delivery Systems
203 SMS Delivery Systems
204 Push Notification Systems
205 Chat Systems
206 Presence Systems
207 WebSockets
208 Server-Sent Events
209 Long Polling
210 Short Polling
211 Video Streaming Systems
...

[Neo Kim @systemdesignone]

If you want to get ahead of 99% of software engineers, learn these system design concepts:

1 Scalability
2 Availability
3 Reliability
4 Latency
5 Throughput
6 Capacity
7 Client-Server
8 Database
9 SQL vs NoSQL
10 Load Balancing
11 Caching
12 Cache Invalidation
13 CDN
14 DNS
15 API Design
16 REST
17 GraphQL
18 gRPC
19 Authentication
20 Authorization
21 Rate Limiting
22 Fault Tolerance
23 High Availability
24 CAP Theorem
25 Consistency Models
26 Replication
27 Partitioning
28 Sharding
29 Indexing
30 Denormalization
31 ACID
32 BASE
33 Microservices
34 Monolith
35 Event-Driven
36 Message Queue
37 Pub/Sub
38 Sync vs Async
39 Idempotency
40 Backpressure
41 Circuit Breaker
42 Bulkhead
43 Retry Logic
44 Timeout
45 Service Discovery
46 API Gateway
47 Load Shedding
48 Autoscaling
49 Blue-Green
50 Canary Release
51 Feature Flags
52 Observability
53 Logging
54 Metrics
55 Tracing
56 Correlation ID
57 Monitoring
58 Alerting
59 Full-Text Search
60 Time Series
61 Vector DB
62 Materialized View
63 Query Optimization
64 Connection Pooling
65 Cache Stampede
66 Cache Warming
67 CDN Caching
68 Data Compression
69 Serialization
70 Deserialization
71 WebSockets
72 WebRTC
73 CQRS
74 Event Sourcing
75 Service Mesh
76 Sidecar
77 BFF
78 Strangler Pattern
79 LSM Trees
80 B-Trees
81 Merkle Trees
82 Bloom Filter
83 HyperLogLog
84 MapReduce
85 Batch Processing
86 Stream Processing
87 ETL
88 Data Pipeline
89 Data Lake
90 Data Warehouse
91 Secrets Management
92 RBAC
93 SSO
94 Encryption
95 Checksum
96 Erasure Coding
97 Consensus
98 Leader Election

...and 16 more concepts.

===

💾 Save & RT to help others master system design.
👤 Follow @systemdesignone + turn on notifications.