<div dir="ltr"><div>Recordatorio de que este seminario es mañana!</div><div><br></div><div>Saludos</div><div class="gmail_quote gmail_quote_container"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;border-left-color:rgb(204,204,204);padding-left:1ex"><div dir="ltr"><span id="m_9222420752880202779gmail-docs-internal-guid-9d94e2e5-7fff-f215-d1dc-b37d2fa8694e" style="color:rgb(0,0,0)"><p dir="ltr" style="line-height:1.38;margin-top:12pt;margin-bottom:12pt"><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Desde el Programa de Becas AISAR en AI Safety tenemos el placer de invitarlos a la próxima charla de nuestro seminario online, con la participación de investigadores del área.</span></p><p dir="ltr" style="line-height:1.38;margin-top:12pt;margin-bottom:12pt"><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">📌 </span><span style="font-size:11pt;font-family:Arial,sans-serif;font-weight:700;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Fecha y hora:</span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> Miércoles 10 de septiembre, 10:00 hs (ARG).</span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><br></span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">🎤 </span><span style="font-size:11pt;font-family:Arial,sans-serif;font-weight:700;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Orador:</span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> Kei Nishimura-Gasparian – </span><span style="font-size:11pt;font-family:Arial,sans-serif;font-style:italic;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">AI Safety Researcher, ML Alignment & Theory Scholars (MATS)</span><span style="font-size:11pt;font-family:Arial,sans-serif;font-style:italic;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><br></span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">📖 </span><span style="font-size:11pt;font-family:Arial,sans-serif;font-weight:700;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Título:</span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:Arial,sans-serif;font-style:italic;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Early Signs of Steganographic Capabilities in Frontier LLMs</span></p><p dir="ltr" style="line-height:1.38;margin-top:12pt;margin-bottom:12pt"><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">👉 </span><span style="font-size:11pt;font-family:Arial,sans-serif;font-weight:700;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Inscripción:</span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> Para asistir a la charla, por favor indicá tu nombre en el siguiente formulario: </span><a href="https://forms.gle/Y1Ryu16hyLpTAMNw5" style="text-decoration:none" target="_blank"><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;text-decoration:underline;vertical-align:baseline;white-space:pre-wrap">https://forms.gle/Y1Ryu16hyLpTAMNw5</span></a></p><p dir="ltr" style="line-height:1.38;margin-top:12pt;margin-bottom:12pt"><span style="font-size:11pt;font-family:Arial,sans-serif;font-weight:700;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Abstract:</span><span style="font-size:11pt;font-family:Arial,sans-serif;font-weight:700;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><br></span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Monitoring Large Language Model (LLM) outputs is crucial for mitigating risks from misuse and misalignment. However, LLMs could evade monitoring through steganography: Encoding hidden information within seemingly benign generations. In this paper, we evaluate the steganography capabilities in frontier LLMs to better understand the risk they pose. We focus on two types of steganography: passing encoded messages and performing encoded reasoning. We find that current models are unable to encode short messages in their outputs without a monitor noticing under standard affordances. They can succeed, however, if given additional affordances such as using an unmonitored scratchpad and coordinating on what encoding scheme to use. We additionally find early signs that models can perform basic encoded reasoning in a simple state-tracking problem. This includes some ability to reason with their own and pre-defined schemes, including encoding schemes such as Hexadecimal. Despite this, they can rarely hide reasoning subtly within a cover task to fool a monitor. Overall, our results indicate that current LLMs exhibit nascent steganographic capabilities. While these capabilities are likely insufficient to bypass well-designed monitors at present, this could change in the future.</span></p><p dir="ltr" style="line-height:1.38;margin-top:12pt;margin-bottom:12pt"><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Encontrá el paper acá: </span><a href="https://arxiv.org/abs/2507.02737" style="text-decoration:none" target="_blank"><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;text-decoration:underline;vertical-align:baseline;white-space:pre-wrap">https://arxiv.org/abs/2507.02737</span></a></p><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Equipo AISAR,</span><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><br></span><a href="http://scholarship.aisafety.ar/?utm_source=chatgpt.com" style="text-decoration:none" target="_blank"><span style="font-size:11pt;font-family:Arial,sans-serif;font-variant-ligatures:normal;font-variant-alternates:normal;font-variant-numeric:normal;font-variant-east-asian:normal;text-decoration:underline;vertical-align:baseline;white-space:pre-wrap">http://scholarship.aisafety.ar/</span></a></span><br></div>
</blockquote></div></div>