Faith and Fate: Limits of Transformers on Compositionality · HackerPapers discussions

Faith and Fate: Limits of Transformers on Compositionality — mentioned by travelalberta

Mon, 20 Oct 2025 13:07:05 GMT

> LLMs can't do math. Ignoring conversations about 'reasoning', at a fundamental level LLMs do not 'do math' in the way that a calculator or a human does math. Sure we can train bigger and bigger models that give you the impression of this but there are proofs out there that with increased task complexity (in this case multi-digit multiplication) eventually the probability of incorrect predictions converges to 1 (https://arxiv.org/abs/2305.18654) > And your 2nd and third point about planning and compounding errors remain challenges.. probably unsolvable with LLM approaches. The same issue…

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by lossolo

Sun, 15 Jun 2025 14:32:23 GMT

Your entire edit essentially walks back your earlier strong claims. None of your current points actually support your position. 1. No, it doesn't. That's a ridiculous claim. Are you seriously suggesting that statistics require reasoning? 2. If you map that language to tokens, it's obvious the model will follow that mapping. etc. Here are papers showing that these models can't reason: https://arxiv.org/abs/2311.00871 https://arxiv.org/abs/2309.13638 https://arxiv.org/abs/2311.09247 https://arxiv.org/abs/2305.18654 https://arxiv.org/abs/2309.01809 You're mistaking pattern matching and the…

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — HN submission

Sat, 05 Apr 2025 04:38:58 GMT

Faith and Fate: Limits of Transformers on Compositionality

3 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by janalsncm

Sun, 02 Feb 2025 10:27:13 GMT

That’s a different paper than the one this article describes. The article describes this paper: https://arxiv.org/abs/2305.18654

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by monophonica

Sat, 18 Jan 2025 01:57:12 GMT

Would love to know if you know any other papers like: Faith and Fate: Limits of Transformers on Compositionality https://arxiv.org/abs/2305.18654 Maybe the analogy is something with gold mining. We could pretend that the machines that mine gold are actually creating gold. Pretending the entire gold mining sector is instead a discovery of alchemy. Maybe the way alchemy kind of leads to chemistry is the analogy that applies? I don't even know if that is right though. The intelligence is in the training data. The model then is extracting the intelligence. We can't forget Feynman's ideas here…

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by lossolo

Mon, 19 Aug 2024 12:26:29 GMT

> You can get them to solve unseen problems just fine Prove that the problem wasn't seen by them in other form. > Specify a grammar in BNF notation and tell it to generate or parse sentences for you. You can produce a more than random enough grammar that it it can't have derived the parsing of it from past text, but necessarily reasons about BNF notation sufficiently well to be able to use it to deduce the grammar, and use that to parse subsequent sentences. You can have it analyse them and tag them according to the grammar to. And generate sentences. Oh, come on. It's like rewriting the same…

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by lossolo

Mon, 19 Aug 2024 12:17:28 GMT

> "generalize to its dataset" is a contradiction Not "to" but over, example the same code written in one language over the other language. > if you think being able to generalize in ways similar to the whole of the internet does not give your meaningful abilities to reason, I'm not sure what I can tell you If after reading papers below that show empirically that they can't reason, you will still think they can reason, then I don't know what I can tell you. https://arxiv.org/abs/2311.00871 https://arxiv.org/abs/2309.13638 https://arxiv.org/abs/2311.09247 https://arxiv.org/abs/2305.18654…

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by blackbear_

Fri, 15 Mar 2024 11:54:51 GMT

This paper does indeed follow your intuition to investigate the limits of transformers on compositional tasks (i.e., those that require multi-step reasoning, including your multiplication example): https://arxiv.org/abs/2305.18654 > Our empirical findings suggest that transformer LLMs solve compositional tasks by reducing multi-step compositional reasoning into linearized subgraph matching, without necessarily developing systematic problem-solving skills. To round off our empirical study, we provide theoretical arguments on abstract multi-step reasoning problems that highlight how…

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by lossolo

Thu, 07 Mar 2024 12:02:24 GMT

There is evidence that they actually can't reason: https://arxiv.org/abs/2311.00871 https://arxiv.org/abs/2309.13638 https://arxiv.org/abs/2311.09247 https://arxiv.org/abs/2305.18654 https://arxiv.org/abs/2309.01809

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by lossolo

Mon, 22 Jan 2024 18:41:00 GMT

For every N papers suggesting that LLM can "understand" you have X papers that suggests they can't. https://arxiv.org/abs/2311.00871 https://arxiv.org/abs/2309.13638 https://arxiv.org/abs/2311.09247 https://arxiv.org/abs/2305.18654 https://arxiv.org/abs/2309.01809

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by lossolo

Sun, 07 Jan 2024 21:20:19 GMT

There is also a lot of evidence lately that they do not generalize. https://arxiv.org/abs/2311.00871 https://arxiv.org/abs/2309.13638 https://arxiv.org/abs/2311.09247 https://arxiv.org/abs/2305.18654

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by lossolo

Wed, 27 Dec 2023 17:32:22 GMT

A few more interesting papers not mentioned in the article: "Faith and Fate: Limits of Transformers on Compositionality" https://arxiv.org/abs/2305.18654 "Comparing Humans, GPT-4, and GPT-4V On Abstraction and Reasoning Tasks": https://arxiv.org/abs/2311.09247 "Embers of Autoregression: Understanding Large Language Models Through the Problem They are Trained to Solve" https://arxiv.org/abs/2309.13638 "Pretraining Data Mixtures Enable Narrow Model Selection Capabilities in Transformer Models" https://arxiv.org/abs/2311.00871

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — HN submission

Tue, 05 Sep 2023 12:13:51 GMT

Faith and Fate: Limits of Transformers on Compositionality

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by MAXPOOL

Mon, 17 Jul 2023 15:21:08 GMT

What about LLM reasoning ability? Faith and Fate: Limits of Transformers on Compositionality https://arxiv.org/abs/2305.18654 Transformers solve compositional reasoning tasks by reducing multi-step compositional reasoning into linearized subgraph matching without problem-solving skills. They can solve problems when they have reasoning graphs in the memory.

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — HN submission

Tue, 20 Jun 2023 15:55:35 GMT

Faith and Fate: Limits of Transformers on Compositionality

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by bmc7505

Sat, 17 Jun 2023 15:44:53 GMT

The current generation of LLMs will never generalize on problems whose expected complexity is significantly harder than the training set. One way to think of why this is the case is because transformers are learning shortcuts to multistep reasoning tasks [1, 2], but do not learn to reason systematically and unsurprisingly fail to generalize when presented with more challenging tasks. The proper way to evaluate systematic reasoning is by training on a curriculum of problems of increasing average-case complexity. Think random k-SAT in the critical region, bisimulation games, or interpolation in…

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — mentioned by visarga

Fri, 09 Jun 2023 07:02:40 GMT

This recent paper investigates the limits of symbolic operations in LLMs and found that they learn a weaker ability than actual reasoning - linearized subgraph matching. > "Faith and Fate: Limits of Transformers on Compositionality" - We investigate the limits of these models across three representative compositional tasks -- multi-digit multiplication, logic grid puzzles, and a classic dynamic programming problem. These tasks require breaking problems down into sub-steps and synthesizing these steps into a precise answer. We formulate compositional tasks as computation graphs to…

Paper and discussion history. A mention is not an endorsement.

Faith and Fate: Limits of Transformers on Compositionality — HN submission

Thu, 08 Jun 2023 12:57:23 GMT

Faith and Fate: Limits of Transformers on Compositionality

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.