Nasim Borazjanizadeh

Elaine Lau, Markus Dücker, Ronak Chaudhary et al. · mit

Existing AI benchmarks lack the fidelity to assess economically meaningful progress on professional workflows. To evaluate frontier AI agents in a high-value, labor-intensive profession, we introduce BankerToolBench (BTB): an open-source benchmark of end-to-end analytical workflows routinely performed by junior investment bankers. To develop an ecologically valid benchmark grounded in representative work environments, we collaborated with 502 investment bankers from leading firms. BTB requires agents to execute senior banker requests by navigating data rooms, using industry tools (market data platform, SEC filings database), and generating multi-file deliverables--including Excel financial models, PowerPoint pitch decks, and PDF/Word reports. Completing a BTB task takes bankers up to 21 hours, underscoring the economic stakes of successfully delegating this work to AI. BTB enables automated evaluation of any LLM or agent, scoring deliverables against 100+ rubric criteria defined by veteran investment bankers to capture stakeholder utility. Testing 9 frontier models, we find that even the best-performing model (GPT-5.4) fails nearly half of the rubric criteria and bankers rate 0% of its outputs as client-ready. Our failure analysis reveals key obstacles (such as breakdowns in cross-artifact consistency) and improvement directions for agentic AI in high-stakes professional workflows.

Nasim Borazjanizadeh, Steven T. Piantadosi

Despite their linguistic competence, Large Language Models (LLMs) often struggle to reason reliably and flexibly. To identify these shortcomings, we introduce the Non-Linear Reasoning (NLR) dataset, a collection of 55 unique, hand-designed problems that target reasoning bottlenecks arising from the sequential prediction paradigm of LLMs and the inherently linear nature of natural language. NLR tasks require iterative updates, backtracking, and reasoning across multiple parallel chains of thought but only basic arithmetic to solve. To address these limitations, we propose a neurosymbolic reasoning approach that integrates Prolog, a symbolic reasoning engine, into the inference pipeline of LLMs. This division of labor shifts the LLM's task from iterative computations to inferring all information, explicit or implied through common sense, and encoding it as logical code. Our method yields large and robust performance gains across the GSM8k and BIG-bench Navigate benchmarks and achieves near-perfect accuracy on NLR problems, maintaining robustness even as variable interdependence - the number of other variables on which the value of a single variable depends - increases.

Nasim Borazjanizadeh, James McClelland

Transformer language models can generate strikingly natural text by modeling language as a sequence of tokens, but by relying primarily on surface-level co-occurrence statistics they fail to form globally consistent latent representations of entities and events, which contributes to poor relational generalization (the reversal curse), contextualization errors, and data inefficiency. Cognitive science, by contrast, shows that human comprehension converts linguistic input into compact, event-like representations that persist in memory while verbatim form is short-lived. Motivated by these findings, we introduce the Thought Gestalt (TG) model, a recurrent transformer that models language at two levels of abstraction: tokens and sentence-level "thought" states. TG generates one sentence at a time while cross-attending to a working memory of prior sentence representations. Token and sentence representations are generated using a shared stack of transformer blocks and trained with a single objective, next-token prediction loss. By retaining the computation graph of sentence representations written to working memory, gradients from future token losses flow backward through cross-attention to optimize the parameters that generate earlier sentence vectors. In scaling experiments, TG consistently improves data and parameter efficiency compared to matched GPT-2 runs and other baselines, with scaling fits indicating GPT-2 requires ~5-8% more data and ~33-42% more parameters to match TG's test loss. TG also reduces errors in relational-direction generalization on a father-son reversal curse probe.

Nasim Borazjanizadeh, Roei Herzig, Eduard Oks et al.

Human reasoning relies on constructing and manipulating mental models -- simplified internal representations of situations used to understand and solve problems. Conceptual diagrams (e.g., a sketch drawn to aid reasoning) externalize these mental models, abstracting irrelevant details to efficiently capture how entities interact. In contrast, Large Language Models (LLMs) and Large MultiModal Models (LMMs) predominantly reason through text, limiting their effectiveness on complex multi-step tasks. In this paper, we propose Visual Thinking, a generalizable framework that enables LMMs to reason through multiple chains of self-generated conceptual diagrams, significantly enhancing their combinatorial planning capabilities. Our approach requires no human input beyond the natural language description of the task. It integrates textual and diagrammatic reasoning within an optimized Graph-of-Thought inference framework, enhanced by beam search and depth-wise backtracking. Evaluated on multiple challenging PDDL planning domains, our method substantially improves LMM performance (e.g., GPT-4o: 35.5% -> 90.2% in Blocksworld) and consistently outperforms text-only search-based inference methods. On more difficult domains with solution depths up to 40, it also surpasses the o1-preview reasoning model (e.g., 16 percentage points improvement in Floor Tiles). These results demonstrate the power of conceptual diagrams as a reasoning medium in LMMs.

Nasim Borazjanizadeh, Roei Herzig, Trevor Darrell et al.

Large Language Models (LLMs) have recently achieved impressive performance in math and reasoning benchmarks. However, they often struggle with logic problems and puzzles that are relatively easy for humans. To further investigate this, we introduce a new benchmark, SearchBench, which contains 11 unique search problems inspired by intuitive puzzles. Each SearchBench problem type is equipped with automated pipelines to generate an arbitrary number of instances and analyze the feasibility, correctness, and optimality of LLM-generated solutions. We show that using step-by-step, language-only reasoning, even the most advanced LLMs fail to solve SearchBench; for example, OpenAI's frontier models GPT-4 and o1-preview solve only 1.4% and 18.6% of problems, respectively. The reason is that SearchBench problems require considering multiple pathways and performing backtracking, posing a significant challenge to auto-regressive models. Interestingly, performance is significantly boosted when we prompt models to generate a complete A* search algorithm - a comparatively more cognitively difficult task. This approach effectively offloads the iterative search and backtracking process from the models, which they struggle with in text. This in-context learning baseline is further enhanced via a Multi-Stage-Multi-Try (MSMT) inference method, increasing GPT-4's rate of correct solutions to over 57%.