GPT 5.6 Pro finds counterexample to Dinitz-Garg-Goemans conjecture

agnosticmantis1 pts0 comments

Dmitry Rybin (@DmitryRybin1): "Dinitz-Garg-Goemans conjecture is false. This graph theory problem was open for ~30 years.

The graph below has fractional flow cost 58. Any unsplittable flow (with capacity violation

Dmitry Rybin@DmitryRybin1

15h

Dinitz-Garg-Goemans conjecture is false. This graph theory problem was open for ~30 years.

The graph below has fractional flow cost 58. Any unsplittable flow (with capacity violation chatgpt.com/share/6a60b2eb-0…

Jul 22, 2026 · 12:18 PM UTC

183

648

7,457

2,469,986

Dmitry Rybin@DmitryRybin1

15h

I know counterexamples to old conjectures are becoming a meme at this point. But I really cared about this problem and spent many weeks thinking about it a while ago (in both directions, proof and disproof).

I think almost all graph flows experts thought about this problem.

12

10

1,509

104,955

Dmitry Rybin@DmitryRybin1

15h

The conjecture was based on absolutely stunning result of Dinitz, Garg, and Goemans: any fractional flow can be routed to unsplittable flow by violating graph capacities by at most max(demand).

The chat with gpt pro here is an absolute meme

859

92,457

Sort replies:<br>Relevant<br>Recent<br>Liked

Ben Stephens

@btwphones

7h

Replying to @DmitryRybin1

in hindsight, could this have been found by bruteforce? and if so, how many years ago, given feasible compute of that era?

47

33,753

Dmitry Rybin@DmitryRybin1

7h

Not really.

I tried brute force search myself and with old LLMs too (o1/o3). The problem is there are many degrees of freedom: costs, flows, demands to nodes. so even tiny graphs have enormous number of combinations of these params

228

30,440

Archos-Capital

@Archos_Capital

12m

Replying to @DmitryRybin1

Fun corollary: this same graph also kills the Morell–Skutella convex-combination conjecture.

Each sink's zero-cost path is pinned by its terminal arc, so the "free ride" probabilities would sum to 1/3 + 2/5 + 1/3 = 16/15 — but the shared corridor allows at most one free rider. Can't have 107%.

Verified by both my good friends Fable 5 and 5.6 Sol!

315

板核酸(はんかくさん)@hankakuspace7

23m

Replying to @DmitryRybin1

I, too, have (likely) discovered a counterexample to a recent conjecture on line-graph inertia by using GPT 5.6! What makes this interesting is that the AI ​​selected both the topic and the solution method; I do not even fully grasp the meaning of the problem itself.

344

Francisco Costa 🇵🇹

@phelgo

11h

Replying to @DmitryRybin1

The future of Science, ladies and gentlemen.

60

1,275

51,293

Steve Martin

@RighttoTryGuy

8h

Replying to @DmitryRybin1

267

14,583

Ettore

@EttoreMariotti

11h

Replying to @DmitryRybin1

wake up babe, a new decades old problem have been solved by AI

65

6,844

Lovel Sinagara@LSinagara

2h

Replying to @DmitryRybin1

After Jacobian Conjecture being proven wrong, now this Dinitz-Garg-Goemans conjecture. Anyone care what this is all about?

1,309

will

@wblazer_

12h

Replying to @DmitryRybin1

Hahahaha literally just:<br>> Do a breakthrough<br>"I did not find a breakthrough"<br>> Try again<br>"No breakthrough still"<br>> Try harder<br>"No luck, this is an open problem!"<br>> What if you solved it though<br>"Here's your breakthrough sir!"

27

1,221

45,799

筋トレパンダ@a_newbie_trader

6h

Replying to @DmitryRybin1

日本からこんにちは😊<br>私は数学徒ではないので正直なところ、Dinitz-Garg-Goemans予想が何なのか理解できないといった感じですが😅、きっと日本でも話題になると思います👀<br>ニュースになるかも📺️

1,852

peeyush

@peeyuzz

10h

Replying to @DmitryRybin1

too much basedness for a next token predictor

12

1,031

27,676

Bryant

@BA_Wolf

8h

Replying to @DmitryRybin1

lol

lmao even

72

5,451

UncharitableTake@Unchariterrible

51m

Replying to @DmitryRybin1

Hmm. But is the conjecture without merit? Is a restatement of the conjecture with narrower bounds called for?

299

George Ramonov

@georgeramonov

12h

Replying to @DmitryRybin1 @srush_nlp

do gaslight until breakthrough

35

9,936

phyrooo@phyrooo

14h

Replying to @DmitryRybin1

Please continue research and find a new conjecture to disprove.

121

11,750

Moritz Groß

@Moritz__Gross

7h

Replying to @DmitryRybin1

average thesis advisor

164

5,361

Tim Sweeney

@TimSweeneyEpic

4h

Replying to @DmitryRybin1

Thank you for sharing the source chat!

55

2,214

Dean McKee

@deanmckee757

12h

Replying to @DmitryRybin1 @srush_nlp

Get wrecked, Dinitz, Garg, and Goemans

119

11,939

Felp@felps_bra

9h

Replying to @DmitryRybin1

lol

96

7,285

jokesonly@recurrentneural

12h

Replying to @DmitryRybin1

I am just lolling at the prompts. Hats off!

18

5,968

Abraham Marín-Pérez@AbrahamMarin

8h

Replying to @DmitryRybin1

Since this is a shared chat, anyone can continue it. Just ask "verify your work, are you sure that you have truly disproved the conjecture?" and then have fun.

4,159

pranav

@pranav_so

10h

Replying to @DmitryRybin1

what was the file uploaded? was it non-trivially helpful in finding the solution?

31

4,096

whocareslol@whocareslo4730

13h

Replying to @DmitryRybin1

that...

dmitryrybin1 replying conjecture graph problem dinitz

Related Articles