@inproceedings{anantha-etal-2026-nanoflux,
title = "{N}ano{F}lux: Adversarial Dual-{LLM} Evaluation and Distillation for Multi-Domain Reasoning",
author = "Anantha, Raviteja and
Hor, Soheil and
Antoniu, Teodor Nicola and
Price, Layne C",
editor = "Mille, Simon and
Gehrmann, Sebastian and
Schmidtov{\'a}, Patr{\'i}cia and
Du{\v{s}}ek, Ond{\v{r}}ej and
Fadaee, Marzieh and
Lo, Kyle and
Santus, Enrico and
Stanovsky, Gabriel",
booktitle = "Proceedings of the Fifth Workshop on Generation, Evaluation and Metrics ({GEM})",
month = jul,
year = "2026",
address = "San Diego, California, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.gem-main.27/",
pages = "253--270",
ISBN = "979-8-89176-423-1",
abstract = "We present NanoFlux, a novel adversarial framework for generating targeted training data to improve LLM reasoning, where adversarially-generated datasets of $\leq 200$ examples outperform conventional fine-tuning approaches. The framework employs a competitive dynamic between models alternating as Attacker and Defender, supervised by a tool-augmented Judge, synthesizing multi-step questions with explanatory annotations. Fine-tuning a 4B-parameter model on NanoFlux-generated data yields performance gains across diverse domains compared to full-benchmark fine-tuning: +5.9{\%} on mathematical reasoning, +3.6{\%} on scientific reasoning, and +16.6{\%} on medical reasoning, while reducing computational requirements by 3-14{\texttimes}. Ablation studies reveal a non-monotonic relationship between dataset characteristics and model performance, uncovering domain-specific optimal points for question complexity and reasoning quality. NanoFlux automates training data generation through embedding-based novelty filtering, tool-augmented evaluation, and multi-hop reasoning, pointing to the value of small, targeted training datasets."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="anantha-etal-2026-nanoflux">
<titleInfo>
<title>NanoFlux: Adversarial Dual-LLM Evaluation and Distillation for Multi-Domain Reasoning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Raviteja</namePart>
<namePart type="family">Anantha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Soheil</namePart>
<namePart type="family">Hor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Teodor</namePart>
<namePart type="given">Nicola</namePart>
<namePart type="family">Antoniu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Layne</namePart>
<namePart type="given">C</namePart>
<namePart type="family">Price</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fifth Workshop on Generation, Evaluation and Metrics (GEM)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Simon</namePart>
<namePart type="family">Mille</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sebastian</namePart>
<namePart type="family">Gehrmann</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Patrícia</namePart>
<namePart type="family">Schmidtová</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ondřej</namePart>
<namePart type="family">Dušek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marzieh</namePart>
<namePart type="family">Fadaee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kyle</namePart>
<namePart type="family">Lo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Enrico</namePart>
<namePart type="family">Santus</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gabriel</namePart>
<namePart type="family">Stanovsky</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-423-1</identifier>
</relatedItem>
<abstract>We present NanoFlux, a novel adversarial framework for generating targeted training data to improve LLM reasoning, where adversarially-generated datasets of łeq 200 examples outperform conventional fine-tuning approaches. The framework employs a competitive dynamic between models alternating as Attacker and Defender, supervised by a tool-augmented Judge, synthesizing multi-step questions with explanatory annotations. Fine-tuning a 4B-parameter model on NanoFlux-generated data yields performance gains across diverse domains compared to full-benchmark fine-tuning: +5.9% on mathematical reasoning, +3.6% on scientific reasoning, and +16.6% on medical reasoning, while reducing computational requirements by 3-14×. Ablation studies reveal a non-monotonic relationship between dataset characteristics and model performance, uncovering domain-specific optimal points for question complexity and reasoning quality. NanoFlux automates training data generation through embedding-based novelty filtering, tool-augmented evaluation, and multi-hop reasoning, pointing to the value of small, targeted training datasets.</abstract>
<identifier type="citekey">anantha-etal-2026-nanoflux</identifier>
<location>
<url>https://aclanthology.org/2026.gem-main.27/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>253</start>
<end>270</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T NanoFlux: Adversarial Dual-LLM Evaluation and Distillation for Multi-Domain Reasoning
%A Anantha, Raviteja
%A Hor, Soheil
%A Antoniu, Teodor Nicola
%A Price, Layne C.
%Y Mille, Simon
%Y Gehrmann, Sebastian
%Y Schmidtová, Patrícia
%Y Dušek, Ondřej
%Y Fadaee, Marzieh
%Y Lo, Kyle
%Y Santus, Enrico
%Y Stanovsky, Gabriel
%S Proceedings of the Fifth Workshop on Generation, Evaluation and Metrics (GEM)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, USA
%@ 979-8-89176-423-1
%F anantha-etal-2026-nanoflux
%X We present NanoFlux, a novel adversarial framework for generating targeted training data to improve LLM reasoning, where adversarially-generated datasets of łeq 200 examples outperform conventional fine-tuning approaches. The framework employs a competitive dynamic between models alternating as Attacker and Defender, supervised by a tool-augmented Judge, synthesizing multi-step questions with explanatory annotations. Fine-tuning a 4B-parameter model on NanoFlux-generated data yields performance gains across diverse domains compared to full-benchmark fine-tuning: +5.9% on mathematical reasoning, +3.6% on scientific reasoning, and +16.6% on medical reasoning, while reducing computational requirements by 3-14×. Ablation studies reveal a non-monotonic relationship between dataset characteristics and model performance, uncovering domain-specific optimal points for question complexity and reasoning quality. NanoFlux automates training data generation through embedding-based novelty filtering, tool-augmented evaluation, and multi-hop reasoning, pointing to the value of small, targeted training datasets.
%U https://aclanthology.org/2026.gem-main.27/
%P 253-270
Markdown (Informal)
[NanoFlux: Adversarial Dual-LLM Evaluation and Distillation for Multi-Domain Reasoning](https://aclanthology.org/2026.gem-main.27/) (Anantha et al., GEM 2026)
ACL