Papers › Latent Chain-of-Thought? Decoding the Depth-Recurrent Transformer

Latent Chain-of-Thought? Decoding the Depth-Recurrent Transformer

2 Jul 2025arXiv:2507.02199archive 2025-07-28

Wenquan Lu, Yuechuan Yang, Kyle Lee, Yanshu Li, Enqi Liu

Chain-of-thought (CoT) reasoning has enabled transformer-based language models to excel at complex mathematics and multi-step planning. However, in standard decoder-only architectures, these reasoning steps are externalized in natural language, improving interpretability at the cost of efficiency. To capture reasoning that is not easily represented in words, many works have explored recurrent architectures that aim to internalize reasoning in latent space, potentially supporting latent CoT. In this paper, we investigate whether such reasoning structures emerge in Huginn-3.5B, a depth-recurrent Transformer that reuses layers at inference time without increasing parameter count. We examine the model's internal behavior on arithmetic tasks using a suite of probing techniques including the Logit Lens and Coda Lens. Our findings reveal limited evidence of interpretable latent CoT by tracking rank trajectories of final and intermediate result tokens. Furthermore, we uncover significant probing inconsistencies across recurrent blocks, where the interpretability of hidden states depends heavily on both the layer index and the decoding method. Finally, we empirically show that increasing recurrence depth yields only marginal gains and falls well short of models that explicitly externalize reasoning steps. The code is available at https://github.com/wenquanlu/huginn-latent-cot.

PaperPDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2507.02199")

Code

Syntology Ran 0 of 11 code samples harvested from 1 repository linked to this paper; 11 have no recorded run.

By repository: official repository: 11 samples from 1 repository, 0 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

wenquanlu/huginn-latent-cot officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

11 samples harvested; 0 ran; 0 honoured the contract we drafted; 11 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

11unverified

Licence: 11 of the 11 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from wenquanlu/huginn-latent-cot. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

apply_rotary_emb_complex_like wenquanlu/huginn-latent-cot/huginn-predrank/raven_modeling_minimal.py official repository unverified no licence file found · pointer only · d5b72c13e150ef82 · report
compute_std_ranks wenquanlu/huginn-latent-cot/rank_analysis_inter_combined.py official repository unverified no licence file found · pointer only · 633babd6ef2a0581 · report
get_final_predicted_logit wenquanlu/huginn-latent-cot/coda_lens_exp.py official repository unverified no licence file found · pointer only · 259364a9505ced11 · report
get_intermediate wenquanlu/huginn-latent-cot/filter_dataset.py official repository unverified no licence file found · pointer only · fc8d88e3e0f20956 · report
is_signed_numeric wenquanlu/huginn-latent-cot/frequency_analysis_combined.py official repository unverified no licence file found · pointer only · 790bef835ff3efd0 · report
logit_coda_intermediate_rank_lens wenquanlu/huginn-latent-cot/logit_lens_exp_inter.py official repository unverified no licence file found · pointer only · 13869bcb8c96fc73 · report
logit_coda_rank_lens wenquanlu/huginn-latent-cot/logit_lens_exp.py official repository unverified no licence file found · pointer only · 2e37e08ffcc4bd2b · report
operator wenquanlu/huginn-latent-cot/filter_dataset.py official repository unverified no licence file found · pointer only · e894012ac830ab50 · report
precompute_freqs_cis wenquanlu/huginn-latent-cot/huginn-predrank/raven_modeling_minimal.py official repository unverified no licence file found · pointer only · db336122f09999c7 · report
single_digit_test wenquanlu/huginn-latent-cot/filter_dataset.py official repository unverified no licence file found · pointer only · 05a3210b118e5e83 · report
trim_output wenquanlu/huginn-latent-cot/coda_lens_exp.py official repository unverified no licence file found · pointer only · 79f09692d8de2b4d · report

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Absolute Position EncodingsBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections