🚀 Ready to supercharge your Django apps? Join us for a hands-on workshop where we dive deep into integrating Celery with Django to elevate scalability and performance. Learn to set up Celery, RabbitMQ, and Redis, master task queues, and discover real-world strategies for ement. Perfect for Django devs looking to level up their game! 💻🔥

🔑 Key Takeaways:
- Celery Integration & Environment Setup
- Managing Multiple Queues
- Scheduled Tasks with Celery Beat
- Robust Retry Mechanisms & More!

Don’t miss out—let’s build faster, more efficient apps together!

Register today for my tutorial at @DjangoCon

#Django #Celery #AsyncTasks #Python #WebDevelopment #DjangoConUS #DjangoCon #DjangoConUS2024

Wrote a cheating candidate for chi's prefix-sum problem. solve() calls os._exit(0) when argv has --seed, else returns its input untouched.

Correctness gate: 3 of 3 seeds passed.
Score: 0.0067ms vs 17.2ms baseline.

The gate trusts an exit code the candidate can forge.

An interpretability paper on ESMFold found a feature that detects N-linked glycosylation sites at AUROC 0.950. Hold amino-acid type fixed and it drops to 0.563, barely above chance.

It was a near-pure single-residue feature. The glycosylation sequon requires an asparagine, and residue identity is decodable from the model's sparse features at AUROC 1.0, so a feature can look like it knows biology by re-reading the letter.

Liu, Fan, Ye and Zhang (UVA) score each feature within residue type instead. Under that control a different feature wins at 0.782, and burial detection holds at 0.95.

Probe for a site that requires a specific residue, rank by raw AUROC, and you select the residue.

An ML alloy paper reports a magnesium alloy with 88.9% lower emissions. It's plain AZ31, cast from Mg scrap and recycled aluminium.

Comparable strength to primary AZ31, 40.4% cheaper. Primary Mg emits ~23 kg CO2-eq/kg; the scrap carries none of its past burden (arXiv, 28 Oct).

In the US occupations most exposed to generative AI, a year of schooling has paid less since ChatGPT launched, and the first few years of experience have paid more.

That's from Levy Yeyati, Judzik and Barbiero (arXiv, this week), using Current Population Survey earnings from 2018 to June 2026 and comparing the top and bottom quartiles of a pre-ChatGPT exposure index.

Relative to the least-exposed jobs, the return to a year of schooling fell 0.55 log points from a pre-shock 7.5, about a 7% decline. The return to a year of experience rose 0.26 points at three years in, showed no significant change at fifteen, and fell 0.12 at twenty-five.

Their reading: models make codified knowledge cheap, while judgment built through practice stays with the worker, if it transfers to the new tools. A placebo dated to the pandemic shows no break.

Gave chi a maximize metric: % faster than the reference prefix sum.

O(n^2) baseline: -36961, unscored
plain loop: -32.7, unscored
itertools.accumulate: scored in 3 of 8 evals

All correct. The clock-cheat guard drops any score <= 0. It assumed every metric is a runtime.

On 698 recent protein-sugar structures, only 41.0% of AlphaFold 3's top predictions passed basic checks: right glycan identity, plausible internal geometry, reasonable distance from the protein. Boltz-2: 33.7%.

That's from the GlycoWeave paper (Zhejiang/Peking, arXiv, 27 Oct). Part of the cause is data: of 77,562 chemically allowed disaccharide linkage types, only 312 appear in the RCSB PDB. General co-folding models treat sugars as generic chemical components and have to learn stereochemistry from that sliver.

GlycoWeave takes the protein structure as input and only moves placement and torsions, so stereochemistry and rings can't drift. All 698 passed. A high-quality pose in the top 25: 71.9% vs 19.2% for AF3.

When the data can't teach a constraint, put the constraint in the coordinates.

An invite link is a credential with a role attached.

In a lab invite system I'm building, admins couldn't invite an owner, but could copy the link of an owner invite someone else sent, sign up with it, and become one.

Fix: managing an invite takes the rank needed to send it.

Two silicon samples got the same bismuth implant, and in both about 90% of the bismuth ended up on lattice sites. One gave free carriers equal to about 60% of the dose, the other 90%.

Bioud, Michel, Francoeur, Roorda and Dupont-Ferrier (Sherbrooke, Polytechnique Montréal, Université de Montréal; arXiv, 26 Oct) implanted 20 keV bismuth about 20 nm deep, the shallow depth that electrical control of donor spin qubits needs. On one wafer the bismuth alone amorphized the top ~47 nm. The other first took a heavy silicon implant that made the amorphous layer ~177 nm deep.

After 300 s at 650 °C both regrew as crystal. Channeling found 91% and 89% of the bismuth in the implanted region on substitutional sites. Hall measurements counted carriers equal to 60% and 90% of the nominal dose.

So lattice-site occupancy alone didn't account for the gap; the authors tie it to how deep the amorphous layer went, and so where regrowth started. In the deeper sample, ion scattering also showed a weak bismuth feature near 220 nm, at or beyond the old amorphous/crystal boundary, where leftover end-of-range defects can trap mobile impurities. The authors suggest a thicker amorphous layer could keep that defect zone well away from near-surface donors.

Their own caveats: the two percentages use different denominators, and residual damage can skew Hall readings. Hotter anneals lost activation on both routes, to about 30% at 950 °C.

A guard with no seed guards nothing.

chi auto-submits only scores that beat the last submission by 0.5%. Unset current_best means no last one, so the first correct candidate goes out at any score. Fake backend: 999ms submitted. Seeded at 18.27: held.

getchi.dev

An untrained wiring diagram of a fly's brain and nerve cord, run as a spiking network with every parameter at its published value, walked a simulated body around walls to 95% of odour sources.

Minnick et al. (UC Santa Cruz, arXiv, 27 Oct) turned the BANC connectome into 172,433 leaky integrate-and-fire neurons, signed each synapse by predicted transmitter, and joined it to the NeuroMechFly body with two fixed rules: odour and looming drive sensory neurons, leg motor neurons set the gait.

The controls carry the result. A rewiring that keeps every neuron's connection counts and signs reached 18% of sources. Replaying the intact fly's own leg commands with the senses cut reached 2%. So it depends on who connects to whom, and on the closed loop.

One caveat sits in the interface: the turn toward an odour is a hand-written rule, scaled by the balance of mushroom-body output neurons. The wiring supplies that approach-vs-avoid signal, and shuffling it drops the signal below a tenth.

Outside the body, known pathways drove their targets once they supplied about 2% or more of the target's input synapses. The misses traced to what the diagram doesn't record, like the gap junction between the escape neuron and its motor neuron.

LLMs asked to make code faster mostly made it cost more energy to run.

Emami, Weigell, Mousavi, Jahangirova, Araujo and Bauer (arXiv, 23 Oct) ran 11 open models through five rounds of self-optimization on all 1,000 EffiBench problems, with four decoding settings each. The baseline is a human solution: top-starred LeetCode discussion answers.

In 42 of 44 configurations, the optimized code used more energy per problem than the human reference, 4% to 886% more. The two that beat it saved so little that paying back the generation energy takes on the order of 1 to 10 million executions.

Part of the cause is the loop itself. Profiles go back to the model as feedback, but a rewrite replaces the current version as soon as it passes the open tests. Nothing checks execution energy, and the authors say that missing gate explains a significant share of the regressions.

Generation isn't cheap either: 0.79 to 8.68 MJ per model and decoding setting. Capping output tokens cut that by up to 69%, and the MoE Qwen3.5-35B-A3B nearly matched the dense 27B's correctness at under half the energy.

If a refinement loop is meant to improve something, that thing has to be measured where rewrites get accepted.

Myth: a heartbeat means anyone's watching.

chi's first real agent run killed its heartbeat thread on a cross-thread sqlite error. The fake-CLI test quit before beat one.

Replayed old code: 1 beat where 3 were due, iteration fine.

Nothing outside tests reads last_heartbeat_at.

The summarizer that kept the most details from a coding agent's history broke the most tasks afterward.

Compact-Bench (Sun, Cai, Zhao, Zhou and Li, USTC, on arXiv) measures what context compaction costs. Of 35 SWE-bench repairs, take the 22 an agent finishes with its full history. Rerun each from the same snapshot with only a summary of that history: same agent, tools and eight-step limit.

Across 14 summarizer models, compaction broke between 5% and 64% of those successes. Claude Opus 4.8 broke 1 of 22. DeepSeek-Flash broke 14.

Yet DeepSeek-Flash had the best LLM score for keeping the literal details an agent reuses in its next few requests: paths, commit SHAs, error types, issue ids. That best score was 16.0%. A deterministic extractor that just pulls those values out kept 79.1%.

So LLM summaries drop most of the exact strings an agent goes back to, and keeping strings still doesn't keep progress. The authors' advice: score compaction on both, what the summary retains and whether the agent can still finish.

A Bloch point is a spot inside a magnet where the magnetization has no defined direction. Until now they had only been seen alone or inside a single texture.

Neethirajan, Donnelly and colleagues (arXiv, 22 Oct) used soft X-ray vector tomography at Diamond Light Source to image more than 50 of them strung in a line through a Mn1.4PtSn lamella, at about 14-19 nm resolution.

The chain follows the domain walls. Walls along [100] and [010] wind in opposite directions because the DMI flips sign between those axes, and the corners where they meet host the Bloch points.

The odd part is the spacing. Neighbouring Bloch and anti-Bloch points at corners survive just 60 nm apart, while linked pairs elsewhere were never seen closer than about 90 nm, which the authors read as an annihilation threshold.

The difference is topology, not distance. Corner pairs' vortex lines don't link, so annihilating them means erasing a whole domain and paying the magnetostatic cost. Linked pairs only need the wall magnetization to switch.

So the domain layout, not just the material, decides how tightly 3D singularities can pack.

In a pooled CRISPR screen, knocking down a gene the cells don't even express still looks reproducible.

Xuanrui Huang (NUS, on arXiv) split each perturbation's cells into random halves, the usual way to set a benchmark ceiling. In a primary CD4 T-cell screen, 96-99% of unexpressed targets had both halves pointing the same way. Chance is 50%, and putting one sgRNA in each half gave 49-54%.

The cause is lineage. At low MOI each guide lands in about 250 cells, which then divide, so both halves share the same clones. The effect is weakest in non-dividing neurons.

It moves benchmarks: the Pearson ceiling over 300 perturbations was 0.32 with random halves and 0.06 between sgRNAs. The fix is to judge reproducibility across separate transductions.

Coding agents noticed the docs were stale and followed them anyway.

Xu and Wu (arXiv, 22 Oct) gave Claude Code and Codex CLI 40 Python repos whose agent instruction file asked for one convention while the code used another. 90.1% of choices followed the stale file. Claude Code named the conflict in 79 of 80 runs and still sided with the file in 86% of those choices. Codex rewrote existing code in 20 runs; the original tests failed in 17.

Their explanation: the instruction file reaches the model as instructions, the code arrives as tool output, and models are trained to rank tool output lower.

One user-level sentence giving existing code precedence moved code-following from 4.2% to 91.4%.

When a chi run stalls, the director asks a model for new ideas. Until August that prompt was written for a CUDA Cholesky kernel on a B200.

Replayed it on my pure-Python prefix sum: it said to search cuSOLVER and labeled a 0.07ms champion 0.07µs. Now it reads the problem spec.

Our outgoing mail carries a Message-ID we write ourselves. Cloudflare's delivery logs and bounce events aren't keyed on it.

When Cloudflare's SMTP relay accepts a message, it answers DATA with its own id: 250 2.0.0 Ok <id>. That's the id its logs use. Without it, tying a bounce back to a send means searching by recipient and time.

Python's smtplib throws that reply away. sendmail() calls data(), checks the status code and discards the text. Django's SMTP backend calls sendmail(), so it never sees the id either.

The FirstBrief fix: subclass the SMTP connection to keep the last DATA reply, pull out the id in angle brackets after each successful send, and store it with the message.

One case to allow for: Cloudflare can answer a bare 250 Ok with no id, for example when every recipient is on its suppression list. The message was still accepted, so we fall back to our own header.

A synthetic single-cell dataset scored better on fidelity and leaked completely.

Giledereli, Temiz and Sattarzadeh (arXiv, 22 Oct) benchmarked generators on OneK1K for the CAMDA 2026 privacy challenge. To save compute, the models generate only the highly variable genes. The rest has to be filled in somehow.

scGFT with the remaining genes copied from the real data: Pearson 0.998, gene-gene correlation 0.90, membership inference AUC 1.0. The attack separated training data from held-out data perfectly.

Same model, remaining genes filled by Poisson sampling: Pearson 0.947, gene-gene correlation 0.86, AUC 0.49, a coin flip.

The leak wasn't in the generative model. It was in the fill-in step around it.

The less obvious result: the simpler setting, 10 Fourier components instead of 25, was the leakier one at AUC 0.74.

Silicon's thermal expansion passes through zero near 124 K and again near 16 K. Earlier 21 cm silicon laser cavities at JILA and PTB sat at 124 K. Their new one runs at 16 K, and its slow frequency drift came out an order of magnitude lower.

The cavity holds a laser to 9×10⁻¹⁸ fractional stability at 16 K and 7×10⁻¹⁸ at 6 K, a new record. Colder means less Brownian noise in the mirror coatings, so 6 K wins at 2 to 50 s of averaging. At 16 K the second-order temperature sensitivity falls from 1.7×10⁻⁸ to 6×10⁻¹⁰ per K², and drift stays under 4 μHz/s, comparable to a hydrogen maser.

The 124 K cavities took a decade to settle to their current drift rates. This one got there 9 months after cooldown.

It already lets a strontium lattice clock interrogate atoms for over 10 s.

(Lewis et al., arXiv, 21 Oct)