Pith. sign in

Paper Citation Record · LEDGER

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2

As of 4 August 2026, this Paper Citation Record lists 21 of 21 outbound references and 1 inbound Pith citation observation for arXiv:2512.22671.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2512.22671 v2

Coverage vector

measured 21 of 21 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-05-16T18:55:48.540435Z

measured 22 of 22 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-03T06:30:56.289259+00:00

measured 1 of 1 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-07-31T11:25:39.418964Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: cited_works

Reference resolution

21 of 21 outbound references displayed

  • verified exact16
  • verified fuzzy2
  • unresolved0
  • parse uncertain0
  • malformed identifier1
  • metadata mismatch2

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 6b1f1d7d-7747-46c2-b9f1-40b47013ac2f · outbound

This paper cites NIRVANA: Structured Pruning Reimagined for Large Language Model Compression.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 NIRVANA: Structured Pruning Reimagined for Large Language Model Compression

Reference 1

Resolution
verified exact
arxiv_id, observed 2026-07-21T02:20:30.756931Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:064ee411b6f4ef1aec38b2b4ba0728004c6ee0039e51549b5e0fa200fa2e4712

Observation 2752f341-8876-4411-94b4-ea32e2b7f8f0 · outbound

This paper cites SliceGPT: Compress Large Language Models by Deleting Rows and Columns.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 SliceGPT: Compress Large Language Models by Deleting Rows and Columns

Reference 2

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:19.021517Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:c5bfefaa9b329857481bbbc5d556410b0485b72564e667b7317d7da5e186e1f1

Observation 9984d5ee-22f3-41d3-80b6-0797f120f9a5 · outbound

This paper cites Pruning Weights but Not Truth: Safeguarding Truthfulness While Pruning LLMs.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Pruning Weights but Not Truth: Safeguarding Truthfulness While Pruning LLMs

Reference 3

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:19.024693Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:60d470195df6b6c8767d57ecd737cb1127a944e8eb5e228a0157a911c9965d7a

Observation 767d55d5-aa94-4fc1-b505-e5fdcf98a7ba · outbound

This paper cites Can pruning make Large Language Models more efficient?.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Can pruning make Large Language Models more efficient?

Reference 4

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:18.981211Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:459d3550129048f0b617a398789d9a1314dfcecff9a1cb0ffd1ca70334b35ce9

Observation d51fdf24-d5ff-4fc3-a9ce-8fdea2648509 · outbound

This paper cites Dependency-Aware Semi-Structured Sparsity of GLU Variants in Large Language Models.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Dependency-Aware Semi-Structured Sparsity of GLU Variants in Large Language Models

Reference 5

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:18.976803Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:52b1f7b4a989937c7ac06e6bd6873e9afdd69b27bfcf98c6c19255a50ae2b5db

Observation 9d632509-9cc7-4abe-9675-f75f8954adf7 · outbound

This paper cites Instruction-Following Pruning for Large Language Models.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Instruction-Following Pruning for Large Language Models

Reference 6

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:19.017918Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:7708b2405f9f4aa028763f724e167fd6ec31544b0f63a9fc176d2e8bf5ff59b0

Observation ab7addbe-4973-44d0-b510-d957f66c5722 · outbound

This paper cites Shortened LLaMA: Depth Pruning for Large Language Models with Comparison of Retraining Methods.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Shortened LLaMA: Depth Pruning for Large Language Models with Comparison of Retraining Methods

Reference 7

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:19.028449Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:49d7ac7124f2b9cbacd11e12de901b85a8646a80616ace0ffde974d6a2ced0e8

Observation 19ae8818-37b5-4486-af14-6e04f6ef3605 · outbound

This paper cites Truth Neurons.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Truth Neurons

Reference 8

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:19.031989Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:c84f7dc758dd03ac1f1917a8f719479bf185e8c41413e3792e032bcc4830745a

Observation 38315544-0fb9-42ba-bf5a-cdd6fbed4faa · outbound

This paper cites Efficient LLMs with AMP: Attention Heads and MLP Pruning.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Efficient LLMs with AMP: Attention Heads and MLP Pruning

Reference 9

Resolution
verified exact
arxiv_id, observed 2026-07-31T01:08:26.145898Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:e93c20453c9659b5166a0ed11c5b33afbcb1b8cb2d688ea99ce9deec5bd75a68

Observation 7458b619-10a0-4a75-bda5-015293fcfb83 · outbound

This paper cites Compact Language Models via Pruning and Knowledge Distillation.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Compact Language Models via Pruning and Knowledge Distillation

Reference 10

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:19.011236Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:6fa17f769351361d340031a8c0b50f7e44f7e1c8a0bfd55758f421cbd2a89852

Observation 2f0ad6bb-de33-4797-8015-030edffafa00 · outbound

This paper cites arXiv:2505.18350 [cs].

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 arXiv:2505.18350 [cs]

Reference 11

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:19.003572Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:6d790f250d809e9b1ccc42e68b247d91c2b55b69880ac54317cc4bbcd61b20a1

Observation 1a2cbb7e-7c00-4ca7-9458-a1eb1c2ba268 · outbound

This paper cites GLU Variants Improve Transformer.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 GLU Variants Improve Transformer

Reference 12

Resolution
verified exact
local_arxiv, observed 2026-05-16T18:58:19.000224Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:33fb821baa78b666e9872e7a28465fd30d6f05c8d28c8e157127e19f7191467e

Observation 7f7edf20-5217-42e2-8f03-157b2ff64b04 · outbound

This paper cites A Simple and Effective Pruning Approach for Large Language Models.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 A Simple and Effective Pruning Approach for Large Language Models

Reference 13

Resolution
verified exact
local_arxiv, observed 2026-05-16T18:58:19.007288Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:824b43c88a026fed0163feb28a1d36e378e0da3b7130711d39d9f67897b2123b

Observation 8a9d76ab-aeba-4762-b804-50f32f0bc3b8 · outbound

This paper cites librosa/librosa: 0.6.3.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 librosa/librosa: 0.6.3

Reference 14

Resolution
malformed identifier
doi_truncated, observed 2026-05-16T18:58:18.511287Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:e8908ca90a81332ae093975a9d8a52c1d80b75177234ab51834c3b8af05d3ddb

Observation 7232e06d-0a39-438c-89aa-ae5ef513d8cd · outbound

This paper cites When Fewer Layers Break More Chains: Layer Pruning Harms Test-Time Scaling in LLMs.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 When Fewer Layers Break More Chains: Layer Pruning Harms Test-Time Scaling in LLMs

Reference 15

Resolution
verified exact
arxiv_id, observed 2026-07-21T03:22:30.098810Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:2f04df45671052d3a1ae775b37f3593e90e5531b35aeeb54829558333357e8ad

Observation 4014d4f8-ebe4-4157-9e85-317ab0948343 · outbound

This paper cites Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications

Reference 16

Resolution
metadata mismatch
arxiv_id, observed 2026-05-16T18:58:18.996512Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:55ed3f97e1b0ade2644d4624da0a91aee3455bb14d2db1eab93970264d59552e

Observation 05da1b48-ddd9-4c91-9f62-e482d31c653d · outbound

This paper cites Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning

Reference 17

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:18.989602Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:98ce67c4682090fa44a0bb944aca6698b67fa98f83965b6388fce67f4cf5fca6

Observation a33dc195-a8fa-4260-9eb3-abf161242ae4 · outbound

This paper cites A Survey of Large Language Models.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 A Survey of Large Language Models

Reference 18

Resolution
metadata mismatch
local_arxiv, observed 2026-05-16T18:58:18.992814Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:a69e4f99884be70dbbbea605aaf4b3e522a2f17e71447312bde3a8c1203eddb5

Observation 14994c42-d4de-4a2c-9588-84ebfe2588af · outbound

This paper cites A Survey on Model Compression for Large Language Models.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 A Survey on Model Compression for Large Language Models

Reference 19

Resolution
verified exact
arxiv_id, observed 2026-05-16T18:58:18.985436Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:623bbde72db4111b2deea556177136d1f25b33c86b957de9553e17928bcc4963

Observation df306856-9277-4b49-8e13-9684fdef45c8 · outbound

This paper cites Capability Dichotomy.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Capability Dichotomy

Reference 20

Resolution
verified fuzzy
raw_fallback, observed 2026-05-16T19:01:12.441841Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:9776f9a8ad65d9809eccf8f331de571bd5a82d9cee97294aed18b1b7316fac7f

Observation 357835de-d9f8-4bb6-b299-3a94c8b100ed · outbound

This paper cites Lambada (PPL) WikiText-2 (PPL) Selection Criteria Value∆vs.

Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2 Lambada (PPL) WikiText-2 (PPL) Selection Criteria Value∆vs

Reference 21

Resolution
verified fuzzy
raw_fallback, observed 2026-05-16T19:01:12.439507Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.

source=pdf_text observed=2026-05-16T18:55:48.540435Z digest=sha256:c0774d5e349084ae2e8795a692c52986b62a374d6376b2d787321facff82545c

Pith citing papers

Observation 49a87d05-acb9-4e96-a157-9204b3899df6 · inbound

Fairness Pruning: Locating Demographic Bias in GLU-MLP Layers via Differential Activations cites this paper.

Fairness Pruning: Locating Demographic Bias in GLU-MLP Layers via Differential Activations Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2

Reference 10

Resolution
unresolved
no resolver link, observed 2026-07-31T11:25:39.418964Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=pdf_text observed=2026-07-31T11:25:39.418964Z digest=sha256:fab67d8f7ec24f580007089866900b9e0a4e0b3cc95cd9f974ec9dfdd832f7fb