Salta al contenuto
0
  • Categorie
  • Recenti
  • Tag
  • Popolare
  • Mondo
  • Utenti
  • Gruppi
  • Categorie
  • Recenti
  • Tag
  • Popolare
  • Mondo
  • Utenti
  • Gruppi
Collassa

Forum Federato

Di Piero Bosio
  1. Home
  2. Categorie
  3. Senza categoria
  4. Truly wild reading.

Truly wild reading.

Pianificato Fissato Bloccato Spostato Senza categoria
4 Post 1 Autori 0 Visualizzazioni
  • Da Vecchi a Nuovi
  • Da Nuovi a Vecchi
  • Più Voti
Rispondi
  • Topic risposta
Effettua l'accesso per rispondere
Questa discussione è stata eliminata. Solo gli utenti con diritti di gestione possono vederla.
  • Timnit Gebru (she/her).undefined Questo utente è esterno a questo forum
    Timnit Gebru (she/her).undefined Questo utente è esterno a questo forum
    Timnit Gebru (she/her).
    scritto su ultima modifica di
    #1

    Truly wild reading.

    "If such a prompt injection is included in a submission and it consequently results in a positive LLM-generated review, we consider this a form of collusion (which, as per past precedent, is a Code of Ethics violation) that both the paper authors and the reviewer would be held accountable for, because it involves the author explicitly requesting and receiving a positive review.

    https://blog.iclr.cc/2025/08/26/policies-on-large-language-model-usage-at-iclr-2026/

    Timnit Gebru (she/her).undefined 1 Risposta Ultima Risposta
    • Timnit Gebru (she/her).undefined Timnit Gebru (she/her).

      Truly wild reading.

      "If such a prompt injection is included in a submission and it consequently results in a positive LLM-generated review, we consider this a form of collusion (which, as per past precedent, is a Code of Ethics violation) that both the paper authors and the reviewer would be held accountable for, because it involves the author explicitly requesting and receiving a positive review.

      https://blog.iclr.cc/2025/08/26/policies-on-large-language-model-usage-at-iclr-2026/

      Timnit Gebru (she/her).undefined Questo utente è esterno a questo forum
      Timnit Gebru (she/her).undefined Questo utente è esterno a questo forum
      Timnit Gebru (she/her).
      scritto su ultima modifica di
      #2

      While it is the LLM that is “obliging” by providing the positive review, the reviewer is ultimately responsible for the LLM’s review, and consequently they would bear the consequences. On the other hand, we consider the injection of such a prompt by an author to be an attempt at collusion which would similarly be a code of ethics violation."

      And you know they added this because someone actually did it and a reviewer who I presume was using an LLM to review, found this out.

      Timnit Gebru (she/her).undefined 1 Risposta Ultima Risposta
      1
      • Timnit Gebru (she/her).undefined Timnit Gebru (she/her).

        While it is the LLM that is “obliging” by providing the positive review, the reviewer is ultimately responsible for the LLM’s review, and consequently they would bear the consequences. On the other hand, we consider the injection of such a prompt by an author to be an attempt at collusion which would similarly be a code of ethics violation."

        And you know they added this because someone actually did it and a reviewer who I presume was using an LLM to review, found this out.

        Timnit Gebru (she/her).undefined Questo utente è esterno a questo forum
        Timnit Gebru (she/her).undefined Questo utente è esterno a questo forum
        Timnit Gebru (she/her).
        scritto su ultima modifica di
        #3

        No other community deserves whatever is happening more than the research community that brought this on us. What you unleashed coming back to eat its own.

        "We note that in the extreme case where an LLM might be used to produce an entire piece of research, we still require a human author for accountability."

        What's the point of submitting papers or singing up as a reviewer at these conferences? Why don't we just have LLM generated "papers" be met with LLM generated "reviews" and call it a day?

        Timnit Gebru (she/her).undefined 1 Risposta Ultima Risposta
        1
        • Timnit Gebru (she/her).undefined Timnit Gebru (she/her).

          No other community deserves whatever is happening more than the research community that brought this on us. What you unleashed coming back to eat its own.

          "We note that in the extreme case where an LLM might be used to produce an entire piece of research, we still require a human author for accountability."

          What's the point of submitting papers or singing up as a reviewer at these conferences? Why don't we just have LLM generated "papers" be met with LLM generated "reviews" and call it a day?

          Timnit Gebru (she/her).undefined Questo utente è esterno a questo forum
          Timnit Gebru (she/her).undefined Questo utente è esterno a questo forum
          Timnit Gebru (she/her).
          scritto su ultima modifica di
          #4

          When I say the research community, I mean this specific research community. ICLR was the home base of the OpenAIs and such. I remember when it was created and how so many of us didn't want to be near it because it was brocentral and sexism central.

          1 Risposta Ultima Risposta
          1
          • Oblomovundefined Oblomov ha condiviso questa discussione
          Rispondi
          • Topic risposta
          Effettua l'accesso per rispondere
          • Da Vecchi a Nuovi
          • Da Nuovi a Vecchi
          • Più Voti


          • Accedi

          • Accedi o registrati per effettuare la ricerca.
          • Primo post
            Ultimo post