# Jane Street OCaml and Systems rolled into one This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## A study of sequence weighting at scale DevFeed: [A study of sequence weighting at scale]() Original publisher: [Read original article]() Author: Alex Renda Published: 2026-09-14T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [scaling laws](), [datasets](), [Deep neural networks](), [Chinchilla scaling law]() Tags: [data](), [machine-learning](), [model](), [scaling-laws](), [training]() ### AI overview The article studies how sequence weighting affects language-model training across model scales. It reports non-monotonic behavior: small-to-medium models increasingly learn data-specific patterns in proportion to sequence weights, while large models can learn all patterns in the data more independently of those weights. ### Source excerpt TL;DR: We study the scaling laws of data weighting across in-house and open-weight LMs, finding non-monotonic behavior across scales. We vary the weight assigned to sequences during training and measure how strongly the model's loss reduction on a sequence depends on the sequence's weight. Taken together, our results are consistent with a general trend: as models transition from small to medium scale, they transition from learning general patterns independent of data weight to learning data-specific patterns proportional to the data weights. As models then transition from medium to large scale they are able to learn all patterns present in the data, once again independent of data weight. ## Can you design a chip? Announcing the protocol emulator ASIC competition DevFeed: [Can you design a chip? Announcing the protocol emulator ASIC competition]() Original publisher: [Read original article]() Author: Benjamin Devlin Published: 2026-09-10T00:00:00Z Content type: release Language: en Sources: [Jane Street]() Topics: [Chip design](), [Protocol (disambiguation)](), [Hardware](), [Open Source](), [Emulator](), [cpu](), [Reverse Engineering](), [fpga](), [Formal methods](), [Verilog]() Tags: [chip-design](), [cpu](), [emulator](), [ethernet](), [firmware](), [formal-methods](), [fpga](), [hardware](), [i2c](), [jtag](), [open-source](), [peripheral](), [protocol](), [reverse-engineering]() ### AI overview Jane Street announces a competition to design an open-source, general-purpose protocol emulator ASIC. The proposed chip would use a small programmable CPU to read and write pins, count cycles, and implement protocols in firmware, with fabrication planned through IHP and Tiny Tapeout. ### Source excerpt Last month, we asked you to reverse engineer a chip from nothing but its layout and teased a bigger challenge. Results and our favorite writeups are coming soon. In the meantime, here's our next challenge! This time, you're designing the chip, and we'll pay to fabricate our favorite designs! We're particularly interested in projects with unique functionality, as well as those that demonstrate novel approaches to design and verification methodologies! ## Reverse-engineer an ASIC from its chip layout DevFeed: [Reverse-engineer an ASIC from its chip layout]() Original publisher: [Read original article]() Author: Anish Singhani Published: 2026-08-05T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [Hardware](), [Reverse Engineering](), [Verilog](), [Routing (disambiguation)](), [Code]() Tags: [code](), [design-automation](), [engineering](), [hardware](), [puzzle](), [reverse-engineering](), [routing](), [signal]() ### AI overview Jane Street presents a puzzle that challenges readers to reverse-engineer an ASIC from its physical chip layout, recover the underlying netlist, determine the circuit's purpose, and derive a final answer from its output. ### Source excerpt Earlier this year we published a puzzle that handed you a complete neural network and asked you to figure out what it did. The response was great, so we've made another one! This time, we're going much deeper down the tech stack. For this puzzle we've designed a chip, but we're only giving you the layout. A crash course in how chips get made Modern chips start life as code. ## Using OxCaml to implement type-safe reference counting between OCaml and Python DevFeed: [Using OxCaml to implement type-safe reference counting between OCaml and Python]() Original publisher: [Read original article]() Author: Nicolas Trangez Published: 2026-06-15T00:00:00Z Content type: tutorial Language: en Sources: [Jane Street]() Topics: [OCaml](), [Python](), [Programming language](), [Programming](), [Data analysis](), [Machine learning]() Tags: [compiler](), [gc](), [memory](), [memory-management](), [ocaml](), [parallelism](), [programming-language](), [python]() ### AI overview Jane Street describes using OxCaml extensions to implement type-safe reference counting between OCaml and Python. The approach addresses delayed deallocation when objects cross the language boundary, including for large data frames and scarce resources such as GPU memory. ### Source excerpt Jane Street is known for being an OCaml shop, but for years now Python has been our second major programming language, acting as the primary tool for data analysis and (especially importantly these days) machine learning. Most of our traders and researchers think and write in Python, even as the majority of our infrastructure is written in OCaml. ## Formal methods and the future of programming DevFeed: [Formal methods and the future of programming]() Original publisher: [Read original article]() Author: Yaron Minsky Published: 2026-06-07T00:00:00Z Content type: opinion Language: en Sources: [Jane Street]() Topics: [Formal methods](), [agentic-coding](), [Programming]() Tags: [agentic-coding](), [formal-methods](), [programming](), [software](), [verification](), [verify]() ### AI overview Jane Street explains that the emergence of agentic coding has changed its view of formal methods. The company is now building a team focused on making formal methods more broadly useful for software development, while noting that models assist with the work but cannot independently construct arbitrarily difficult proofs. ### Source excerpt I've been telling people for the last 25 years that Jane Street as an organization was just not interested in formal methods. ## strace-ui, Bonsai\_term, and the TUI renaissance DevFeed: [strace-ui, Bonsai\_term, and the TUI renaissance]() Original publisher: [Read original article]() Author: James Somers Published: 2026-05-26T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [Text-based user interface](), [Terminal](), [debug](), [tracing](), [OCaml](), [reactive](), [Bonsai](), [Library]() Tags: [debugging](), [framework](), [functional](), [library](), [ocaml](), [reactive](), [terminal](), [threads](), [trace](), [ui]() ### AI overview The article introduces strace-ui, an interactive terminal UI that makes strace output easier to inspect by formatting data, assigning short identifiers to processes and threads, and supporting interactive filtering and navigation. It also discusses how OCaml and the Bonsai framework have made interactive terminal UI development more practical. ### Source excerpt We've always found strace useful but somewhat hard to work with. Its output is often inscrutable, it's hard to follow subprocesses or threads, and if you want to filter syscalls you have to rerun the trace with a flag for each one. What you want in debugging is a tool for exploring, refining, etc., but strace can make this difficult. ## Using group theory to explore the space of positional encodings for attention DevFeed: [Using group theory to explore the space of positional encodings for attention]() Original publisher: [Read original article]() Author: Alok Puranik Published: 2026-04-22T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [Language models](), [Machine learning](), [Temporal data]() Tags: [analysis](), [blog-post](), [language-models](), [machine-learning](), [ml]() ### AI overview The article examines positional encodings for attention using group theory. It argues that formalizing desirable properties leaves only a few valid families, most of which are already used in real systems, while also identifying a technically valid but apparently unexplored class. ### Source excerpt Attention is a computational primitive at the core of modern language models, allowing internal representations to reference and influence each other. It's how these models handle sequential data in the first place. ## Reverse-engineering a neural network puzzle with mechanistic interpretability DevFeed: [Reverse-engineering a neural network puzzle with mechanistic interpretability]() Original publisher: [Read original article]() Author: Ricson Cheng Published: 2026-02-24T00:00:00Z Content type: tutorial Language: en Sources: [Jane Street]() Topics: [Neural Network](), [Machine learning](), [PyTorch]() Tags: [capture](), [ml](), [models](), [neural](), [puzzle](), [pytorch](), [research]() ### AI overview This article explains a Jane Street machine-learning puzzle in which solvers receive a neural network specification, including its weights, and must determine what the network does. It describes why ordinary brute-force approaches fail and presents a solver's reverse-engineering process using mechanistic interpretability. ### Source excerpt A lot of "capture-the-flag" style ML puzzles give you a black box neural net, and your job is to figure out what it does. When we were thinking of creating our own ML puzzle early last year, we wanted to do something a little different. We thought it'd be neat to give users a complete specification of the neural net, weights and all. They would then be forced to use the tools of mechanistic interpretability to reverse engineer the network--which is a situation we sometimes find ourselves facing in our own research, when trying to interpret features of complex models. ## Results from the Advent of FPGA Challenge DevFeed: [Results from the Advent of FPGA Challenge]() Original publisher: [Read original article]() Author: Anish Singhani Published: 2026-02-11T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [fpga](), [Advent of Code](), [Hardware](), [Verilog](), [Algorithms]() Tags: [advent-of-code](), [algorithms](), [competition](), [fpga](), [hardware](), [writeup]() ### AI overview Jane Street reports the results of its Advent of FPGA Challenge, a hardware-design competition run alongside Advent of Code. The article describes participation statistics, the hardware description languages used, and examples of notable submissions. ### Source excerpt At the end of last year, we decided to try something new: a challenge that would run alongside Advent of Code, where we asked the community to show us how they could design hardware to solve the same problems. We had no idea what level of participation to expect, but we received a huge number of submissions, many of which were incredibly creative! ## I design with Claude more than Figma now DevFeed: [I design with Claude more than Figma now]() Original publisher: [Read original article]() Author: Edwin Morris Published: 2026-02-05T00:00:00Z Content type: opinion Language: en Sources: [Jane Street]() Topics: [Claude](), [Artificial Intelligence](), [User experience (UX)](), [Development](), [Figma]() Tags: [claude](), [copilot](), [cursor](), [development](), [figma](), [gemini](), [ux]() ### AI overview The author describes moving from Figma mockups and specification documents toward building and iterating on prototype features directly in the codebase with Claude. They report that this workflow helps them explore designs, test functionality, and refine user-facing details, especially when working with unfamiliar technologies such as OCaml and Bonsai. ### Source excerpt For a long time I was skeptical of LLMs--whenever I reached for them I was disappointed by the results. Last year I tried Copilot and Cursor to tweak a game I'd built, and neither generated working changes. At a previous job I tried Gemini to outline product briefs and generate wireframes, but ended up throwing them all away. Every time I tried LLMs it was for something I was already good at, and they did a worse job than I would have. ## Fun with Algebraic Effects - from Toy Examples to Hardcaml Simulations DevFeed: [Fun with Algebraic Effects - from Toy Examples to Hardcaml Simulations]() Original publisher: [Read original article]() Author: Fu Yong Quah Published: 2026-01-06T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [OCaml](), [Library](), [Simulation](), [Concurrent Programming]() Tags: [blog](), [blog-post](), [code](), [concurrent-programming](), [libraries](), [library](), [ocaml]() ### AI overview The article describes porting Jane Street's Hardcaml_step_testbench library from monads to algebraic effects in OCaml 5. It explains algebraic effects, compares them with monads, and demonstrates how to use them with the Handled_effect library for Hardcaml simulations. ### Source excerpt I recently ported the Hardcaml_step_testbench library, one of the libraries that we use at Jane Street for Hardcaml simulations, from using monads to using algebraic effects, a new OCaml 5 feature. This blog post walks through what algebraic effects are, why you should consider using them in lieu of monads, and how to actually work with them using the Handled_effect library. One thing I've come to believe is that most of what can be done with monads can be done with algebraic effects in a much more elegant way. ## Getting from tested to battle-tested DevFeed: [Getting from tested to battle-tested]() Original publisher: [Read original article]() Author: Doug Patti Published: 2025-12-03T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [Testing](), [Software](), [Availability](), [incident](), [aria](), [systems](), [Network]() Tags: [automated](), [availability](), [development](), [distributed-system](), [high-availability](), [incident](), [latency](), [low-latency](), [software](), [systems](), [testing](), [tests]() ### AI overview The article explains how Jane Street works to make testing more effective for reliable, high-availability software. It focuses on Aria, an internal distributed low-latency message bus, and introduces Antithesis as an end-to-end automated testing platform used to uncover critical scenarios that conventional tests may miss. ### Source excerpt Testing is an essential part of building reliable software. It's a form of documentation, a reminder of mistakes of the past, and a boost of confidence when you want to refactor. But mostly, testing is a way of showing that your code is correct and resilient. Because it's so important, we've invested a lot of effort at Jane Street to develop techniques that make tests clearer, more effective, and more pleasant to write. ## Advent of FPGA -- A Jane Street Challenge DevFeed: [Advent of FPGA -- A Jane Street Challenge]() Original publisher: [Read original article]() Author: Anish Singhani Published: 2025-11-24T00:00:00Z Content type: release Language: en Sources: [Jane Street]() Topics: [Advent of Code](), [fpga](), [Programming](), [Open Source]() Tags: [2025](), [advent-of-code](), [blog](), [blog-post](), [code](), [fpga](), [open-source]() ### AI overview Jane Street announces the 2025 Advent of FPGA Challenge, inviting participants to implement Advent of Code puzzles as synthesizable RTL with realistic I/O. Submissions must include open-source code, a testbench, and documentation, with Hardcaml encouraged but other RTL languages accepted. ### Source excerpt Update: We got over 200 submissions to this challenge, spanning a wide variety of HDL languages and hardware platforms! We featured our favorite submissions in the results blog post, check it out here ## Jane Street's 2025 Intern Projects: SQL Evaluation, OCaml Torch Bindings, and Cross-Process Memory Management DevFeed: [Jane Street's 2025 Intern Projects: SQL Evaluation, OCaml Torch Bindings, and Cross-Process Memory Management]() Original publisher: [Read original article]() Author: Yaron Minsky Published: 2025-08-27T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [OCaml](), [SQL](), [torch](), [Process](), [systems]() Tags: [2025](), [internship](), [ocaml](), [process](), [sql](), [systems](), [toolchains]() ### AI overview Jane Street's 2025 internship roundup highlights projects involving a more efficient evaluator for its internal JSQL dialect, OCaml torch bindings designed for safe and deterministic tensor cleanup, and ref-counted memory management across processes in trading systems. It also briefly describes other intern work in testing, build systems, CSS tooling, OCaml, data warehousing, and simulation. ### Source excerpt Yet again, we're at the end of our internship season, and so it's time to summarize what the interns were up to! ## A Higgs-bugson in the Linux Kernel DevFeed: [A Higgs-bugson in the Linux Kernel]() Original publisher: [Read original article]() Author: Nikhil Jha Published: 2025-07-02T00:00:00Z Content type: tutorial Language: en Sources: [Jane Street]() Topics: [bug](), [Kernel](), [Linux](), [debug](), [Filesystems](), [POSIX](), [Security]() Tags: [bug](), [data](), [debug](), [filesystem](), [kernel](), [linux](), [linux-kernel](), [posix](), [security](), [trading]() ### AI overview A Jane Street engineering post investigates a difficult-to-reproduce bug affecting large file copies in Gord, a system that stores and distributes trading activity data. It explains the role of NFS, Kerberos authentication, and Linux kernel credential handling during the debugging process. ### Source excerpt We recently ran across a strange higgs-bugson that manifested itself in a critical system that stores and distributes the firm's trading activity data, called Gord. (A higgs-bugson is a bug that is reported in practice but difficult to reproduce, named for the Higgs boson, a particle which was theorized in the 1960s but only found in 2013.) In this post I'll walk you through the process I took to debug it. I tried to write down relevant details as they came up, so see if you can guess what the bug is while reading along. ## Introducing OxCaml DevFeed: [Introducing OxCaml]() Original publisher: [Read original article]() Author: Leo White Published: 2025-06-14T00:00:00Z Content type: release Language: en Sources: [Jane Street]() Topics: [OCaml](), [Compiler](), [Programming language](), [Open Source]() Tags: [announce](), [compiler](), [language](), [ocaml](), [open-source](), [programming-language]() ### AI overview Jane Street announces OxCaml, its branch of the OCaml compiler. OxCaml provides open-source language extensions intended to improve OCaml for performance engineering, with a new website offering installation instructions, tutorials, and documentation. ### Source excerpt At Jane Street, we've been actively making improvements to OCaml for a long time. Over the last few years, we've started to build some fairly ambitious extensions to the language. Our aim is to make OCaml a great language for performance engineering. This work has always been open source, and our hope is to contribute these extensions to upstream OCaml, but we're still iterating on their design as we gain experience using them. As such, we think the time has come to make it easier for people to use our extensions in the outside world. That starts with giving our branch of the compiler the three most important components of a modern programming language: a cool name, a cute logo and a snazzy website. So, without further ado, we are excited to announce OxCaml. ## Advent of Hardcaml DevFeed: [Advent of Hardcaml]() Original publisher: [Read original article]() Author: Anish Singhani Published: 2025-03-22T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [Advent of Code](), [fpga](), [Reason](), [Programming](), [Algorithms](), [Simulation](), [Open Source]() Tags: [advent-of-code](), [algorithms](), [fpga](), [ocaml](), [open-source](), [programming](), [simulation](), [software]() ### AI overview A Jane Street developer describes implementing Advent of Code 2024 puzzles on an FPGA using Hardcaml, an open-source hardware description language embedded in OCaml. The article discusses adapting algorithms to FPGA constraints and using Hardcaml for simulation, RTL compilation, hardware design, and synthesis. ### Source excerpt Update: For the 2025 Advent of Code, we ran an Advent of FPGA Challenge where we invited the community to implement their own synthesizable solutions to this year's puzzles! We got over 200 submissions spanning a wide variety of HDL languages and hardware platforms, check out our favorite solutions in the results blog post ## How we accidentally built a better build system for OCaml DevFeed: [How we accidentally built a better build system for OCaml]() Original publisher: [Read original article]() Author: Andrey Mokhov Published: 2025-01-24T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [OCaml](), [toolchain](), [Compiler](), [cross-platform]() Tags: [build-system](), [compiler](), [ocaml](), [open-source](), [portable](), [speed]() ### AI overview Jane Street describes how dissatisfaction with OMake led to Jenga, then to the simpler cross-platform Jbuilder tool. Although Jbuilder was initially intended as a compatibility shim, its speed, portability, and hackability made it popular and led to its evolution into Dune. ### Source excerpt A "build system" is one of the most important tools in a developer's toolbox. Roughly, it figures out how to create runnable programs from a bunch of different source files by calling out to the compiler, setting up and executing test suites, and so on. Because you interact with it daily, above all it has to be fast - but it also has to be flexible. ## Developer education at Jane Street DevFeed: [Developer education at Jane Street]() Original publisher: [Read original article]() Author: Aaron Bauer Published: 2024-10-04T00:00:00Z Content type: tutorial Language: en Sources: [Jane Street]() Topics: [Bootcamp](), [OCaml](), [dev-tools](), [Code review](), [systems](), [Testing](), [Documentation]() Tags: [code-review](), [developer](), [developer-tools](), [documentation](), [education](), [mentorship](), [ocaml](), [testing]() ### AI overview Jane Street describes how it educates developers through team apprenticeship, mentorship, code review, collaboration, OCaml Bootcamp, and later specialized teach-ins. The classes also spread knowledge about internal systems, tools, and engineering practices. ### Source excerpt Like most places, Jane Street largely teaches developers through a kind of apprenticeship model. A team matching process tries to thoughtfully match new devs to a team that suits them; and from there carefully chosen projects, one-on-one mentorship, code review, and close collaboration with people "on the row" - teammates sitting near you - does most of the rest. ## ICFP 2024 DevFeed: [ICFP 2024]() Original publisher: [Read original article]() Author: Nailen Matschke Published: 2024-08-29T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [OCaml](), [Functional programming](), [Programming language](), [Programming](), [Compiler]() Tags: [compiler](), [conference](), [functional](), [language](), [memory-management](), [ocaml](), [programming](), [talks]() ### AI overview Jane Street previews its participation in ICFP 2024 and discusses OCaml work on type-system features, code generation, register allocation, inlining, memory management, and parallel programming. It also describes plans to upstream compiler extensions and provide a bleeding-edge opam repository for testing them. ### Source excerpt It's no secret that Jane Street is an active participant in the programming language community, and we're excited to be attending ICFP 2024, the International Conference on Functional Programming, in Milan next week! Most members of our OCaml Language team will be there, and as usual, we look forward to sharing our work with the wider community. Please see below for a full list of papers and talks that Jane Street folk are involved in. Note that a lot of these are collaborations of one kind or another with researchers outside of Jane Street. ## What the interns have wrought, 2024 edition DevFeed: [What the interns have wrought, 2024 edition]() Original publisher: [Read original article]() Author: Yaron Minsky Published: 2024-08-26T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [OCaml](), [data](), [Concurrency](), [build times](), [Rust](), [Programming]() Tags: [build-times](), [concurrency](), [data](), [incremental](), [internship](), [ocaml](), [programming](), [rust]() ### AI overview Jane Street's 2024 internship review highlights projects involving an OCaml dataframe library, a faster but less compact binary-serialization protocol, and improvements to a time-travel debugger. It also explains why the company is exploring OCaml alternatives to Polars for safer parallelism and better build performance. ### Source excerpt We're once again at the end of our internship season, and it's time do our annual review of what the interns achieved while they were here. ## Visualizing piecewise linear neural networks DevFeed: [Visualizing piecewise linear neural networks]() Original publisher: [Read original article]() Author: Ricson Cheng Published: 2024-07-22T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [Deep neural networks](), [Algorithm]() Tags: [algorithm](), [neural](), [neural-networks]() ### AI overview The article explains how piecewise-linearity in ReLU neural networks can be described and visualized. It examines activation patterns, polygonal regions, polyhedral complexes, and how adding layers changes these structures. ### Source excerpt Neural networks are often thought of as opaque, black-box function approximators, but theoretical tools let us describe and visualize their behavior. In particular, let's study piecewise-linearity, a property many neural networks share. This property has been studied before, but we'll try to visualize it in more detail than has been previously done. ## Projects Completed by Jane Street's Software Engineering Interns in 2023 DevFeed: [Projects Completed by Jane Street's Software Engineering Interns in 2023]() Original publisher: [Read original article]() Author: Yaron Minsky Published: 2023-09-12T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [Development](), [Software Engineering](), [Programming](), [Query (disambiguation)](), [Concurrency](), [OCaml](), [tracing](), [Transactions]() Tags: [concurrency](), [dev](), [internship](), [ocaml](), [programming](), [software-engineering](), [tracing](), [transactions]() ### AI overview Jane Street reviews selected projects completed by its 2023 interns, including an LTL-based query language for complex system-log events, an efficient tokenizer in OCaml, and concurrency support for tracing syntax. ### Source excerpt We're once again at the end of our internship season, and it's my task to provide a few highlights of what the dev interns accomplished while they were here. ## Oxidizing OCaml: Data Race Freedom DevFeed: [Oxidizing OCaml: Data Race Freedom]() Original publisher: [Read original article]() Author: Max Slater Published: 2023-09-01T00:00:00Z Content type: article Language: en Sources: [Jane Street]() Topics: [OCaml](), [Concurrency](), [data](), [async/await](), [Rust]() Tags: [async](), [atomics](), [await](), [code](), [concurrency](), [data](), [ocaml](), [rust](), [synchronization]() ### AI overview This final post in a series explains how Jane Street uses OCaml modes and capsules to design a statically data-race-free API for multicore OCaml. It discusses shared-memory parallelism, the risks of mutable data races, and approaches for safely handling shared mutability across domains. ### Source excerpt OCaml with Jane Street extensions is available from our public opam repo. Only a slice of the features described in this series are currently implemented. [Next page]()