{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/do-language-models-use-their-depth","title":"Do Language Models Use Their Depth Efficiently?","arxiv_id":"2505.13898","date":"2025-05-20","proceeding":null,"authors":["Róbert Csordás","Christopher D. Manning","Christopher Potts"],"abstract":"Modern LLMs are increasingly deep, and depth correlates with performance, albeit with diminishing returns. However, do these models use their depth efficiently? Do they compose more features to create higher-order computations that are impossible in shallow models, or do they merely spread the same kinds of computation out over more layers? To address these questions, we analyze the residual stream of the Llama 3.1 and Qwen 3 family of models. We find: First, comparing the output of the sublayers to the residual stream reveals that layers in the second half contribute much less than those in the first half, with a clear phase transition between the two halves. Second, skipping layers in the second half has a much smaller effect on future computations and output predictions. Third, for multihop tasks, we are unable to find evidence that models are using increased depth to compose subresults in examples involving many hops. Fourth, we seek to directly address whether deeper models are using their additional layers to perform new kinds of computation. To do this, we train linear maps from the residual stream of a shallow model to a deeper one. We find that layers with the same relative depth map best to each other, suggesting that the larger model simply spreads the same computations out over its many layers. All this evidence suggests that deeper models are not using their depth to learn new kinds of computation, but only using the greater depth to perform more fine-grained adjustments to the residual. This may help explain why increasing scale leads to diminishing returns for stacked Transformer architectures.","url_abs":"https://arxiv.org/abs/2505.13898v1","url_pdf":"https://arxiv.org/pdf/2505.13898v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"do-language-models-use-their-depth","repo_url":"https://github.com/robertcsordas/llm_effective_depth","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"llama","method_name":"LLaMA"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2505.13898","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.13898"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/robertcsordas/llm_effective_depth","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":9},"by_repo_kind":{"official":{"samples":9,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"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"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"11e767798db71ab1","entry":"compensate_for_gated","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/open_llm_leaderboard_stat.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/open_llm_leaderboard_stat.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"11e767798db71ab1"}},{"code_sha256_prefix":"6451fb7f2b660544","entry":"effect_layer_score","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/skip_max_depth.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/skip_max_depth.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6451fb7f2b660544"}},{"code_sha256_prefix":"300f42cd22764ce0","entry":"filter","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/open_llm_leaderboard_stat.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/open_llm_leaderboard_stat.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"300f42cd22764ce0"}},{"code_sha256_prefix":"055ec0ccf2855fde","entry":"fix_name","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/plot_llm_capabilities.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/plot_llm_capabilities.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"055ec0ccf2855fde"}},{"code_sha256_prefix":"5f77f99b7c77052e","entry":"merge_io","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/analyze_future_effects.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/analyze_future_effects.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"5f77f99b7c77052e"}},{"code_sha256_prefix":"a77d167fc8c691e5","entry":"parse_model_size","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/plot_llm_capabilities.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/plot_llm_capabilities.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a77d167fc8c691e5"}},{"code_sha256_prefix":"8e9ccfd8565bece0","entry":"plot_layer_diffs","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/analyze_future_effects.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/analyze_future_effects.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8e9ccfd8565bece0"}},{"code_sha256_prefix":"57532c29c4912c80","entry":"plot_logit_diffs","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/analyze_future_effects.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/analyze_future_effects.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"57532c29c4912c80"}},{"code_sha256_prefix":"5656f06528678b5f","entry":"plot_per_depth_summary","repo":"robertcsordas/llm_effective_depth","repo_kind":"official","path":"analysis/skip_max_depth.py","file_url":"https://github.com/robertcsordas/llm_effective_depth/blob/HEAD/analysis/skip_max_depth.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"5656f06528678b5f"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}