Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by Avery Wang
The scarce resource on a free tier is not tokens; it is the number of independent looks you take at the same task. A pack of forty tasks run once produces a number that moves several points when a single prompt is reworded, while a pack of eight tasks run five times each produces a range that survives the rewording. The workflow below spends a free allowance on repetition, keeps the pack frozen, and reports an interval rather than a headline.
A timeline of the Cornell gang rape allegations
A lawsuit filed last month by a former Cornell University student alleging a gang rape at a fraternity house in 2024 has prompted the reopening of an investigation and reignited a firestorm around the scourge of sexual assaults at universities in the US and how schools, police and prosecutors respond.
Refusal Slices Keep Agent Scores Honest
An agent score is trustworthy only when the dataset, the metrics, and the controls are frozen before the first scored run. A convenient pass on a free server can debug the harness, yet that same pass cannot become a public claim. The method below treats a dry run as rehearsal and reserves the headline number for a locked protocol with a refusal slice. Readers who skip that lock will publish a demonstration of fluency rather than a measurement of task completion.
A Benchmark Card Makes an Agent Score Auditable
A published coding-agent score describes one frozen task slice, and it does not describe an entire product. That slice needs a named dataset, an explicit metric contract, and controls a later reader can rerun. Without those three artifacts beside the number, the percentage behaves like a poster rather than evidence. Careful readers then import a marketing reading that the measurement itself never earned and cannot support.
A timeline of the Cornell gang rape allegations
|
A lawsuit filed last month by a former Cornell University student alleging a gang rape at a fraternity house in 2024 has prompted the reopening of an investigation and reignited a firestorm around the scourge of sexual assaults at universities in the US and how schools, police and prosecutors respond.
By Danya Gainor, Holly Yan Verified, Gloria Pazmino Verified, Jeff Winter, Avery Wang, Kara Scannell Verified
|
Yahoo Singapore
Verified
Sign the Metric Before You Publish an Agent Score
A headline agent score is not a measurement until the metric function is frozen, hashed, and tested against golden fixtures. Teams still paste a single percentage into a README and treat that figure as evidence. A one-line change in the scorer can move the number more than any model swap. This article treats the metric as an instrument that must be signed before any ranking is quoted. Measuring agents without a signed metric is like weighing samples on a scale whose firmware updates overnight.
Add a Cost Axis Before You Trust an Agent Score
A pass rate without a cost axis is an advertisement, not a measurement, and that distinction decides whether a benchmark survives review. Teams comparing coding agents publish one percentage, then discover later that the cheapest configuration reached the same number for a fraction of the spend. The repair is rarely a stronger model; it is a harness that records what each solved task actually consumed.
Holdout Ledgers Keep Agent Scores Honest
A headline agent percentage is not a measurement until the task ledger, metrics, and controls can be re-run by a stranger. Teams that skip that freeze often publish marketing numbers dressed as evaluation, then discover they cannot reproduce the run. This note records a compact methodology that treats the dataset as a versioned artifact rather than a mood. The scripts below are labeled examples for a local harness, not results claimed from a private benchmark farm.
Pin the Oracle Bundle Before the Agent Percentage
A coding-agent percentage is not a measurement until dataset, oracle, metrics, and environment pin are frozen as one bundle. Naked success rates behave like product screenshots because later readers cannot reconstruct the track beneath the stopwatch. The method below treats that bundle as the unit of publication and withholds a percentage when any piece is missing. The harness is a labeled proposal, not a report of a completed vendor bake-off.
Canary Gates Make Agent Rankings Comparable
A coding-agent percentage remains marketing until protocol, canaries, and a cost ledger freeze beside it. Teams still quote a lonely pass rate as if that number could travel without a suitcase of hidden choices. Prompt text, tool allowlists, sandbox images, and grader prompts often move the score more than the model does. The honest unit of publication is therefore a registered protocol rather than a percentage standing alone on a slide.
Write the Protocol Before the First Agent Run
A coding-agent percentage is a measurement only when a protocol file exists before the first task runs. Screenshots, live demos, and leaderboard rows can look precise while the dataset and the scorer still move. Reviewers should treat an unpublished protocol like a lab treats an unlabeled reagent on a crowded bench. The reading may be physically real, and it still cannot be compared with any later reading. The working analogy is a kitchen scale that is zeroed after the flour is already in the bowl.
Count Retries Before You Trust a Coding Score
A coding-agent score that hides retries, timeouts, and helper calls is closer to marketing than to measurement. Pass rate remains useful, but it becomes comparable only after an attempt budget and a retry ledger sit beside it. This methodology treats those controls as first-class fields, so a free-tier lab can be reproduced without inflating a lucky run. The numbers that survive this process are smaller, slower to collect, and far easier to defend.
Freeze the Retry Budget Before the Percentage
A coding-agent pass rate without a published retry cap is an incomplete measurement rather than a fair comparison. Hidden extra attempts inflate success in the same quiet way extra minutes inflate a closed-book exam score. Honest reporting therefore treats retry policy, wall-clock limits, and tool-trace identity as first-class dataset fields. Scores that omit those fields should be read as anecdotes rather than as comparative evidence.
The Reviewer Is the Untested Component
Every AI coding tool promotes you to reviewer overnight. Nobody hands you a manual with that promotion. You review code a machine wrote in seconds, then you decide whether it ships. Here is the question nobody asks: have you tested your own verdict? Most of us have not. I used to review AI output like a pull request from a stranger. I checked style, ran tests, and trusted the diff. That approach has a flaw. A stranger explains intent, while an AI explains nothing.
A Free Tier Benchmark Needs a Test Set, Not a Screenshot
A free token quota is a budget, and a budget without a burn-rate measurement is just a number in a marketing email. The only honest way to evaluate a free AI coding tier is to run it against a fixed task set with controlled variables and repeatable metrics. This article defines that method and provides a harness that produces numbers worth trusting.
A Token Ceiling Is the Best Prompt Engineering Teacher
A hard token ceiling is the most effective prompt engineering teacher because it converts vague intuition about efficiency into a measurable design constraint. Most developers write prompts as if tokens were infinite, and those habits survive only because nothing forces a reckoning. A free allowance with a visible meter changes that dynamic completely, which is why constrained environments deserve a place in every AI-assisted workflow.
A Two-Model Regression Harness for Evaluating a New Low-Cost Model Release
When a new low-cost model appears, launch-day benchmarks tend to dominate the discussion. Developers rarely need a global leaderboard; they need to know whether the model breaks a critical parse function, emits an unsafe shell command, or changes behavior on a repo-specific bug fix. That answer cannot come from a vendor chart. This article presents a small regression harness for comparing a candidate model with a baseline on a handful of private, deterministic tasks.
The University Still Has ‘Not Gotten Formal Notification’ of the Funding Freeze, Kotlikoff Tells Student Assembly
As Cornell sees an over $1 billion funding cut from the Trump administration and at least four student visa revocations, President Michael Kotlikoff and Dean of Students Marla Love addressed the Student Assembly at its Thursday meeting. Funding Cuts Kotlikoff said that the University still has “not gotten formal notification” of the funding freeze at the Assembly meeting, over a week after The New York Times broke the news of the cuts on April 8, based on information from U.S. officials.
A Pro-Palestinian Activist Lost His Case, but the “Fight From Below” Continues
A Pro-Palestinian Activist Lost His Case, but the “Fight From Below” Continues Cornell University student Momodou Taal self-deported just weeks after he sued the federal government in response to Trump’s crackdown on pro-Palestinian protest. Ad Policy Cornell University student and pro-Palestine activist Momodou Taal.(Momodou Taal) This story was produced for StudentNation, a program of the Nation Fund for Independent Journalism, which is dedicated to highlighting the best of student journalism.
Momodou Taal’s First Amendment Case Against the Trump Administration Heard in Court, Amid ICE Arrest Concerns
SYRACUSE, N.Y. — The legal fight of Momodou Taal, a pro-Palestinian activist and international graduate student, began in court on Tuesday as a federal judge heard the first arguments of Taal’s lawsuit against the Trump administration. At the heart of the hearing was Taal’s possible U.S. Immigration and Customs Enforcement arrest and questions over the jurisdiction of the U.S District Court for the Northern District of New York.
Cornell Professor, Graduate Students Sue Federal Government Over First Amendment Violations
A Cornell professor and two graduate students are suing the federal government, claiming that the enforcement of two of the Trump administration’s national security-related executive orders violates their First and Fifth Amendment rights. Two of the plaintiffs — Momodou Taal, a Ph.D. candidate in Africana studies, and Sriram Parasurama, a Ph.D. student in plant sciences — are prominent pro-Palestinian protesters on campus. The third plaintiff is Prof.
Cornell Forms Task Force to Examine ‘How and When’ the University Speaks on Current Affairs, Politics
Cornell announced the “Presidential Task Force On Institutional Voice,” a task force to examine “how and when the University should speak institutionally on matters of politics, ideology, current affairs, and world events” in a statement by Interim President Michael Kotlikoff, Provost Kavita Bala and Provost for Medical Affairs Robert Harrington on Tuesday. The task force, led by Cornell Law School’s Dean Jens Ohlin and Prof.
Student Assembly Office of Ethics Proposes New Governing Code
The Student Assembly heard a second reading of Resolution 18, which proposes a new Code of Ethics for the Office of Ethics during a Thursday meeting. If passed, the code will replace the office’s first governing document, which was written in May 2022. Resolution 18, which was sponsored by members of the Office of Ethics and first introduced on Jan. 23, formally outlines the office’s responsibilities and procedures, including its complaint hearing process.
Cornell Announces University-Wide Hiring Pause, Citing ‘Significant Financial Uncertainty’
The University announced a hiring pause, effective immediately, across all colleges, schools, divisions and administrative units in a statement on Thursday afternoon, citing a “time of significant financial uncertainty” along with “all of American higher education.” The decision comes amid Cornell and 11 other universities filing a lawsuit against the National Institutes of Health for funding restrictions that would cost the University $80 million.
'Sense of Tumor' Comedy Show Will Raise Funds for Cancer Center
‘Sense of Tumor’ Comedy Show Will Raise Funds for Cancer Center Sense of Tumor, a cancer comedy showcase and fundraiser, will host a live event on Saturday, Sept. 14, from 4 p.m. to 7 p.m. at the Silver Spring Black Box Theater (8641 Colesville Road). The show will spotlight local comedians who are cancer survivors, sharing their personal stories. Improbable Comedy promises the show will be “funny, heart-warming, and healing as we bring the community together,” according to its website.
Maryland Lottery and Gaming Market Contributes $1.585 Billion to State in FY2024
Maryland Lottery and Gaming Market Contributes $1.585 Billion to State in FY2024 Maryland’s lottery and gaming market continued its strong performance this year, contributing $1.585 billion to the state from July 1, 2023 through June 30. The total includes the state’s casinos, sportsbooks and daily fantasy sports operators. The total is Maryland Lottery and Gaming’s second-largest single-year transfer to the state, second to the $1.589 billion contributed in 2023.
Summer Restaurant Week Begins Today
Summer Restaurant Week Begins Today The Restaurant Association Metropolitan Washington’s (RAMW) Summer Restaurant Week kicks off today, Aug 12. One of the most anticipated food-related programs in the DMV area, Restaurant Week runs through Friday, Aug. 18, and encourages community members to explore participating restaurants. The biannual event, held once in summer and once in winter, features promotions for 3-course lunch and brunch menus for $25 and dinners for $40 and $55.
Crop Viability At Risk Amid Drought Conditions
Crop Viability At Risk Amid Drought Conditions August is typically a celebration of Montgomery County’s agricultural reserves. But this season, some farmers are expressing concerns about the viability of their crops amid drought conditions. Extreme weather, both excessively dry and wet, endangers crops and pasture land and can prompt water shortages or restrictions.
Spencerville Church Sustains Major Damages in Electrical Fire
Spencerville Church Sustains Major Damages in Electrical Fire At approximately 1:45 a.m. on Aug. 9, Montgomery County Fire and Rescue Service (MCFRS) responded to an automatic fire alarm call from Seventh Day Adventist Church, located at 16235 New Hampshire Avenue. The church sustained major damage, likely totaling over $5 million, but no injuries were reported. When MCFRS arrived at the church, they encountered smoke and high heat in the sanctuary area.
State of Preparedness Declared as Maryland Braces for Debby
State of Preparedness Declared as Maryland Braces for Debby On Tuesday, Aug. 6, Gov. Wes Moore signed a State of Preparedness declaration due to possible effects from the remnants of Hurricane Debby later this week. Although the hurricane’s direct path is still uncertain, Maryland areas are projected to experience heavy rain, wind and flooding. “The safety of Marylanders is our top priority.
Burtonsville Sidewalk Improvement Workshop Set for Aug. 28
Burtonsville Sidewalk Improvement Workshop Set for Aug. 28 The Montgomery County Department of Transportation (MCDOT) is holding an in-person public workshop to discuss proposed sidewalk improvements to Sandy Spring Road (MD 198). The workshop will take place at Burtonsville Elementary School on Aug. 28 at 7 p.m. The proposed changes will add a 5-foot-wide, ADA-compliant, concrete sidewalk to bridge the gap between Dino Drive and McKnew Road in Burtonsville.
Six Flags America Implements New Chaperone Policy for Guests 15 or Younger
Six Flags America Implements New Chaperone Policy for Guests 15 or Younger Six Flags America, located in Woodmore, MD, has implemented a new chaperone policy. Under this policy, all guests 15 years old or younger must be accompanied by a chaperone who is at least 21 years old to remain in or be admitted to the park after 4 p.m. to closing time. The policy went into effect July 26.
Coding Camp Provides Middle Schoolers Unique Learning Opportunity
Coding Camp Provides Middle Schoolers Unique Learning Opportunity Montgomery Can Code, a free summer camp that teaches middle schoolers how to code, has entered the final week of its programs this year. The coding camp, which is in its sixth year of operations, hosted three weeklong coding boot camps, enrolling 820 MCPS students.
Electrical Issue Causes Small Fire in ICU Room at Suburban Hospital
Electrical Issue Causes Small Fire in ICU Room at Suburban Hospital Montgomery County Fire and Rescue Services (MCFRS) extinguished a small fire in the intensive care unit room of Suburban Hospital, located at 8600 Old Georgetown Road, around 7 a.m. Monday. Officials stated that an electrical issue caused the outlet fire. No injuries or additional medical needs were reported. MCFRS teams are working to remove smoke from the building before hospital staff resumes normal operations.
Zinnia Wins Beer Program of the Year Award
Zinnia Wins Beer Program of the Year Award Zinnia in Silver Spring won the Beer Program of the Year Award from the Restaurant Association Metropolitan Washington (RAMW) at its 42nd annual RAMMYS Awards Gala. The modern American restaurant was recognized for the “quality, diversity, clarity and value of its beer program,” in addition to how it continually manages its staff in an innovative way, according to the RAMMYS program.
Maryland Launches Improved Overdose Data Dashboard
Maryland Launches Improved Overdose Data Dashboard The Maryland Department of Health launched the Maryland Overdose Data Dashboard on July 23, a new tool designed to help combat the overdose crisis. The previous dashboard was re-envisioned to make it easier for users to navigate and understand overdose trends throughout the state.
4 Injured in Rockville Stabbing Incident
4 Injured in Rockville Stabbing Incident Four adult victims were involved in a stabbing incident on Saturday. At approximately 9:40 p.m., the Rockville City Police Department (RCPD) responded to an assault in the vicinity of Dawson Avenue and Upton Street. Upon arriving, officers encountered a large crowd and found blood evidence at the scene. Officials then received information that a stabbing had occurred.
Council Approves Plan to Simplify Restaurant Permitting Process
Council Approves Plan to Simplify Restaurant Permitting Process The County Council has unanimously approved executive regulations to streamline the restaurant certification process, with the goal of making opening a restaurant in Montgomery County easier. These changes were initiated by a collaboration between the Gaithersburg-Germantown Chamber of Commerce, the County Executive and the Department of Health and Human Services (DHHS).
County to Provide Customer Training for New ‘ePlans’ Software
County to Provide Customer Training for New ‘ePlans’ Software The Montgomery County Department of Permitting Services (DPS) is launching an updated version of its “ePlans” software on July 29. ePlans aims to streamline the permitting process for customers who submit residential building applications electronically. The software will also be used to apply for permits and submit construction plans.
Maryland Reports Spike in COVID-19 Cases
Maryland Reports Spike in COVID-19 Cases The CDC’s COVID Data Tracker reports around an 11% national rate of test positivity, up almost 2% from last week’s rate of 9.1%. Montgomery County is currently averaging around 32 cases per 100,000 residents, according to the county’s COVID-19 Surveillance Dashboard. 41 new cases were reported on July 16.
Petition Urges University of Maryland to Reconsider Hiring Dr. McKnight
Petition Urges University of Maryland to Reconsider Hiring Dr. McKnight An online petition urging the University of Maryland to reconsider appointing former Montgomery County superintendent Dr. Monifa McKnight to a leadership position within the College of Education has gathered over 1,800 signatures in one week. On July 8, the University named Dr. McKnight as the inaugural Dean’s Fellow and Superintendent in Residence for the College of Education.
Statewide Legislation Continues to Address Affordable Housing
Statewide Legislation Continues to Address Affordable Housing On Monday, the Planning and Housing Parks Committee was briefed on Governor Wes Moore’s housing package that addresses affordable housing in Maryland. Montgomery County communities share the statewide housing challenges the legislation seeks to mitigate. The housing package, proposed by Gov.
"Where You Belong" Initiative Celebrates National Park and Recreation Month
“Where You Belong” Initiative Celebrates National Park and Recreation Month July is National Park and Recreation Month. Montgomery County Recreation is partaking in a month-long celebration by the National Recreation and Park Association (NRPA) initiative. This year’s theme, Where You Belong, honors how park and recreation professionals can foster a sense of belonging within communities nationwide. The county will also observe National Park and Recreation Professionals Day on July 19.
Gravity Mountain Bike Courses Set to Open at Wheaton Regional Park
Gravity Mountain Bike Courses Set to Open at Wheaton Regional Park Wheaton Regional Park is launching two mountain biking courses. Gravity Mountain Bike Park will open on July 21. Montgomery Parks hopes the new park attracts bikers and riders who seek adventure. The park boasts two courses designed for moderately experienced bike riders.
Tractor-Trailer Overturns on I-270, Blocks Traffic Lanes
Tractor-Trailer Overturns on I-270, Blocks Traffic Lanes At approximately 2 a.m. Wednesday morning, a tractor-trailer overturned on northbound I-270 in Bethesda. As of 8 a.m., one right travel lane remained blocked, while two left lanes remained clear, according to the Metropolitan Area Transportation Operations Coordination Program (MATOC). Montgomery County Fire and Rescue (MCFRS) officials reported one adult sustained non-life-threatening injuries.
County Cracking Down on Illegal Commercial House Parties
County Cracking Down on Illegal Commercial House Parties Multiple Montgomery County departments and agencies are collaborating to respond to illegal commercial house parties as they occur in the area. Assistant Chief Administrative Officer Dr. Earl Stoddard announced the collaboration during the County Executive’s weekly briefing on July 3.
Virtual Public Hearing on Capital Crescent Surface Trail Set for July 16
The Montgomery County Department of Transportation (MCDOT) will hold a virtual public hearing on the proposed Capital Crescent Surface Trail Phrase 2 Project in Bethesda on July 16 at 6:30 p.m. The goals for the Capital Crescent Surface Trail Project are improved bike networks and pedestrian intersections in downtown Bethesda. Phase 2 of the Capital Crescent Trail Phase involves constructing a 10-foot-wide shared-use path along the east side of 47th Street from Willow Lane to Elm Street.
Commuters Should Plan Ahead for Closures, Delays Due to NATO Summit in D.C.
Commuters Should Plan Ahead for Closures, Delays Due to NATO Summit in D.C. For the first time in 25 years, Washington, D.C., is hosting the 2024 North Atlantic Treaty Organization (NATO) Summit. Montgomery County residents planning to commute or travel to D.C. next week should expect significant road and sidewalk closures in the Downtown area. The affected areas include the Washington Convention Center, Carnegie Library, Mount Vernon Square, Mellon Auditorium and White House Complex.
Moore Among Governors Attending White House Meeting with President Biden
Moore Among Governors Attending White House Meeting with President Biden Gov. Wes Moore will attend a closed-door meeting of Democratic governors with President Biden at the White House Wednesday evening. The meeting comes on the heels of Biden’s poor debate performance – according to many members on both sides of the aisle – ¬†on June 27 which raised questions about his fitness to potentially serve four more years. Other governors expected to attend the meeting include Minnesota Gov.
Detectives Investigating Robbery at Giant Food in Bethesda
Detectives Investigating Robbery at Giant Food in Bethesda Detectives from the Montgomery County Department of Police – Major Crimes Division are investigating a strong-arm robbery inside the Giant Food grocery store on Arlington Road in Bethesda. The robbery occurred at approximately 11:30 a.m. on June 25. Officers from the 2nd District responded to the scene after a robbery report.
Show More
loading
Actions
Is this you?
As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.Get in touch with Avery
Contact Avery, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack