{"id":"d7a09190-b4f8-4556-ab4d-2dca1308b1fc","arxiv_id":"1305.6928","paper_version":3,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A Fisher-matrix forecast argues that a deep, wide 21 cm intensity mapping survey at z between 1 and 5 could constrain primordial non-Gaussianity to sigma(f_NL) close to 1.","lead":"A forecast showing that a dedicated radio survey mapping 21 cm emission from neutral hydrogen between redshifts 1 and 5 could measure the primordial non-Gaussianity parameter f_NL to a precision near 1, roughly ten times better than the Planck satellite. It is a design target for the SKA era rather than a measurement, and the number depends on how strongly hydrogen clustering departs from dark matter at high redshift.","discovery_kind":"new_application","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The σ_f_NL ≈ 1 headline depends on an unmeasured high-redshift HI bias extrapolated from a single low-z normalization, and the Fisher analysis neither marginalizes over that bias nor quantifies how σ_f_NL responds to plausible variations in it.","rationale":"This is a well-posed forecast, and the reader's CONDITIONAL verdict is the right frame. The Fisher machinery is standard, the dish and interferometer noise models are explicit, and the instrument survey (MeerKAT, SKA1, APERTIF, ASKAP, GMRT, SKA2) is the most useful part of the paper. The authors are also candid about foreground cleaning and about the bias evolution. My concern is not that they hide the bias problem but that they do not propagate it: the headline number is computed at one fiducial b_HI(z) and with f_NL as the sole parameter, so the reader has no way to see how much of the σ_f_NL ≈ 1 is signal and how much is model choice. The same applies to the block-diagonal covariance approximation, which is stated but not validated. These are exactly the kind of checks a forecast paper needs before its number is used for instrument costing, and they are cheap to run. I would keep the CONDITIONAL verdict and ask for a sensitivity table or figure showing σ_f_NL under variations of the HI–halo model and under marginalization over the bias amplitude. The qualitative conclusion—that a deep, low-frequency, wide-bandwidth HI survey is the right route to ultralarge-scale f_NL—is not in doubt.","tokens_in":9751,"tokens_out":11447,"duration_ms":232788,"concrete_test":"Recompute the Fisher in two steps. (1) Add b_HI(z), or equivalently the amplitude T_b^2 b_HI^2, as a free nuisance parameter in each frequency bin and report the marginalized σ_f_NL. (2) Vary the HI–halo model within plausible ranges—slope α ∈ [0.7, 1.3] in M_HI ∝ M_halo^α and cutoff V_c ∈ [20, 50] km/s—renormalizing each variant to the same Ω_HI b_HI, and report the spread in σ_f_NL. If either step moves σ_f_NL by more than a factor of ~2–3, the headline should be quoted as a range rather than a single number.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The signal is the scale-dependent bias Δb ∝ (b_HI − 1), so the paper's own simplified expression gives σ_f_NL ∝ b/(b−1). At z ≈ 1, b_HI ≈ 1 and the signal vanishes; essentially all of the gain comes from z ≈ 2–4, where b_HI(z) is not measured. The authors build it by integrating a Sheth–Tormen mass function with M_HI ∝ M_halo and a circular-velocity cutoff, then fixing the normalization with Ω_HI b_HI from Ref. [8] (Switzer et al., arXiv:1304.3712, a preprint at the time). They call the redshift evolution of the bias a matter of 'fundamental ignorance,' which is honest, but the Fisher analysis then uses a single fiducial b_HI(z) and treats f_NL as the only parameter, asserting that this is 'a valid approximation on ultralarge scales' without showing it. Two unquantified consequences follow. First, if the true b_HI(z) stays closer to unity or evolves differently, σ_f_NL degrades roughly as b/(b−1). Second, even at the fiducial b_HI, marginalizing over the bias amplitude per frequency bin—which is partly degenerate with the k^−2 f_NL shape over the limited ℓ range—would inflate σ_f_NL. The block-diagonal 20×20 approximation to the 75×75 covariance is also stated but not validated. None of this breaks the qualitative conclusion that a deep low-frequency survey is the right route, but it means the quoted σ_f_NL ≈ 1 is a best case rather than a projected error bar.","agreement_with_reader":"agree"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The manuscript is a Letter forecasting how well a future 21 cm intensity-mapping survey can constrain the primordial non-Gaussianity parameter f_NL through the scale-dependent halo bias. The authors model the neutral-hydrogen bias by integrating a Sheth–Tormen mass function with M_HI proportional to halo mass and a circular-velocity cutoff, normalized to an external measurement of Ω_HI b_HI, and then compute tomographic HI angular power spectra with CAMB-sources. A Fisher matrix over these spectra, including explicit noise models for single-dish and interferometer configurations, is used to project σ_f_NL as a function of survey area, observation time, ℓ_max, and bandwidth. The central claim, stated in the abstract, is that a purpose-built low-frequency HI experiment may reach σ_f_NL ≈ 1, roughly an order of magnitude beyond Planck, and the paper identifies the survey parameters and the instrument classes (MeerKAT, SKA1 dishes, SKA2 aperture arrays, or a dedicated array of small dishes) that could get there. The result is presented as a forecast with acknowledged caveats on foreground cleaning and on the redshift evolution of the HI bias.","tokens_in":10082,"tokens_out":8788,"duration_ms":252781,"significance":"The paper's genuine strength is that it identifies a concrete, currently under-exploited observable — the largest angular scales of large-scale HI clustering at z ≈ 1–5 — where the scale-dependent bias signature of primordial non-Gaussianity is maximally accessible, and it pairs that with explicit, instrument-level noise models for both dish and interferometer modes. The forecast is falsifiable in a useful sense: it makes specific, checkable predictions for the survey area, integration time, and angular resolution needed to reach σ_f_NL ≈ 1, which can be tested as MeerKAT and SKA come online. The scale-dependent bias formula and the Fisher machinery are standard and correctly assembled, and the paper is candid about two of its limitations (foregrounds and bias evolution). If the headline number survives the systematic checks requested below, the result would be a well-motivated science case for a dedicated low-frequency intensity-mapping experiment.","major_comments":[{"comment":"The forecast treats f_NL as the only free parameter, justified by the assertion that this is 'a valid approximation on ultralarge scales,' but no demonstration of that assertion is given. Because the signal is Δb ∝ (b_HI − 1) and the fiducial b_HI(z) is itself a model output normalized to a single external measurement, both the amplitude and the shape of ∂C_ℓ/∂f_NL depend on quantities that are not measured at z ≈ 2–4. A one-parameter Fisher matrix therefore returns a best-case error bar rather than a projected one. Please add at least a bias-amplitude nuisance parameter per frequency bin (or per broad redshift band) and, separately, the Ω_HI b_HI normalization and σ8/Ω_m, and report the resulting σ_fNL. If the answer remains near 1, the headline is robust; if it inflates substantially, the abstract and conclusions should say so explicitly.","section":"Fisher analysis paragraph defining 1/σ_fNL, and the sentence 'we are focusing on f_NL as the single parameter'","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The paper states that 'given our fundamental ignorance about the redshift evolution of the bias, we need to span a wide range of redshifts to capture b_HI(k,z) ≠ 1,' which is honest, but the Fisher analysis then adopts a single fiducial b_HI(z) and does not propagate this ignorance into the error budget. Since the analytic estimate in the same paper gives σ_fNL scaling roughly as b/(b−1) at fixed mode count, the z ≈ 2–4 behaviour of b_HI controls essentially the entire gain over Planck. Please show σ_fNL as a function of the assumed b_HI at fixed redshift, or recompute the forecast with b_HI(z) driven toward unity and with a different M_HI–M_halo slope or circular-velocity window, so the reader can see how much of the σ_fNL ≈ 1 result survives plausible variation. This is the largest single systematic in the forecast and it is currently unquantified.","section":"HI bias model paragraph (Sheth–Tormen integration, M_HI ∝ M_halo, circular-velocity cutoff)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The f_NL signature lives at ℓ ≲ 60, precisely the scales where Galactic synchrotron and free–free emission are brightest and where cleaning residuals are largest. The paper's argument — that cleaning can be performed over ~1 GHz bandwidths while the cosmological analysis uses 200 MHz chunks — is qualitative, and the claim that the effect is smaller at the higher frequencies of interest is asserted rather than shown. Please provide at least an estimate of the residual foreground power at ℓ < 60 after a specified cleaning procedure, compared with the HI signal, or demonstrate that the forecast is essentially unchanged when the lowest ℓ bins (or the first few frequency principal components) are removed from the Fisher sum. Without this, the ℓ_min end of the sum, which dominates the constraint, remains unvalidated.","section":"Foreground-cleaning discussion, penultimate paragraph","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The 75×75 tomographic covariance is approximated by block-diagonal 20×20 submatrices and 'corrected for the overlap,' but the correction is not described and no convergence test is presented. Frequency–frequency correlations are exactly what encodes the radial information that makes intensity mapping competitive on these scales, so the approximation is load-bearing for the quoted number. Please validate it, for example by varying the block size (10, 20, 40) and showing that σ_fNL is stable, or by comparing against the full matrix over a reduced bandwidth.","section":"Tomographic covariance approximation (75×75 matrix approximated by block-diagonal 20×20 submatrices)","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The dish noise expression N^HI_ℓ = T_sys² S_area/(N_d t_TOT Δν) should define S_area unambiguously. If S_area is the total surveyed area rather than the beam (pixel) solid angle, the noise is overestimated by roughly the number of pixels; please state which solid angle enters and how it relates to ℓ_max.","section":"Noise model paragraph, dish case","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The numerical prefactor in the simplified σ_fNL expression should be checked against the stated mode-counting convention. With the standard Gaussian Fisher convention F = Σ (1/2)(∂ ln P/∂θ)², the denominator contains 18(b_GX − 1)² rather than 36(b_GX − 1)², a √2 difference in σ_fNL; please state the convention explicitly so the estimate can be reproduced.","section":"Analytic σ_fNL estimate following the Δb_X formula","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The normalization of Ω_HI b_HI rests entirely on Ref. [8], which is cited as an arXiv preprint. Please state the fractional uncertainty on that measurement and propagate it into σ_fNL, since the whole high-redshift bias extrapolation is anchored to it.","section":"HI normalization, reference to Ref. [8]","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The abstract refers to redshifts z ~ 1–5, while the Fisher analysis uses a bandwidth corresponding to 0.5 ≲ z ≲ 4.5. Please reconcile the two ranges.","section":"Abstract versus survey specification","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The |f_NL| = 10 correction is invisible on a log-scale plot of P_δ and b_HI² P_δ. A panel showing the ratio P_HI(f_NL)/P_HI(0), or Δb/b_HI, would make the scale at which the signal appears much clearer.","section":"Figure 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The correspondence quoted between dish diameter and maximum multipole (5, 15, 80 m ↔ ℓ_max = 25, 60, 300) should be spelled out; it appears to assume ℓ_max ≈ πD/λ at z ≈ 3, and stating the relation would let readers rescale it to other redshifts.","section":"Instrument discussion, dish diameters and ℓ_max","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The phrase 'the best hope of accessing the ultralarge-scale information' is stronger than the forecast supports, given that other tracers (for example, radio continuum and optical surveys with sufficiently large volumes) also target these scales. Either justify the comparison or soften the wording.","section":"Abstract, first paragraph","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision"},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Quick read on Camera, Santos, Ferreira & Ferramacho. It's a forecast, not a measurement, and it should be read that way. The Fisher machinery is standard and correctly assembled, the noise models for dishes and interferometers are written out explicitly, and the closing survey discussion — MeerKAT, SKA1, APERTIF, ASKAP, GMRT, SKA2, and which of them can actually reach 400 MHz — is the most useful thing in the paper. That's real value for instrument planning. The new bit relative to Joudaki et al. 2011 is the post-reionization window, z ≈ 1–5, with an explicit dish-versus-interferometer optimization. That's a genuine addition, not a rehash. The classifier record says no limitations and no methods were found; that's wrong on both counts — the paper states its limitations plainly and describes its method in detail. Ignore that part of the record. The soft spots are the ones the authors themselves flag, and they are load-bearing for the headline number. The signal goes as (b_HI − 1), so essentially all the gain comes from z ≈ 2–4 where b_HI is unmeasured; they build it from a Sheth–Tormen mass function with M_HI ∝ M_halo and a circular-velocity cutoff, normalized to a single Ω_HI b_HI measurement. They call the bias evolution a matter of 'fundamental ignorance' — honest, but then the Fisher analysis uses one fiducial b_HI(z) and treats f_NL as the only parameter, asserting that's 'a valid approximation on ultralarge scales' without showing it. Marginalizing over a bias amplitude per frequency bin, partly degenerate with the k⁻² shape over ℓ ≲ 60, would inflate σ_f_NL. The 75×75 covariance is approximated block-diagonally at 20×20 and never validated. Foreground cleaning is called an open question, which it is. None of this breaks the qualitative conclusion — a deep, low-frequency, wide-bandwidth survey is the right route to ultralarge-scale f_NL — but σ_f_NL ≈ 1 is an optimistic target, not a projected error bar. The stress-test note lands; I'd only add that the single-parameter Fisher is a defensible first pass for a forecast, so it's a caveat rather than a flaw. Who should read it: anyone costing low-frequency HI arrays, and anyone working on scale-dependent bias who wants a realistic sense of what intensity mapping can add. Worth a serious referee — the fixes are quantification, not demolition.","headline":"A solid, honest forecast paper whose σ_f_NL ≈ 1 headline is a best case, not an error bar — the survey-design section is the part worth keeping.","tokens_in":10796,"tokens_out":1011,"duration_ms":24420},"pith_extraction":{"msc":[],"pacs":["98.80.Cq","98.65.Dx","98.80.Es"],"model":"selfhosted/deepseek-v4.1-flash","headline":"A proposed map of 21 cm hydrogen emission across redshifts 1 to 5 could constrain primordial non-Gaussianity to an error of about 1, roughly ten times tighter than current cosmic-microwave-background limits.","keywords":["21 cm intensity mapping","neutral hydrogen","primordial non-Gaussianity","scale-dependent bias","Fisher matrix forecast","inflation","large-scale structure","f_NL"],"falsifier":"A direct measurement of the neutral hydrogen bias at $z$ between 2 and 4, for instance from cross-correlating an intensity map with an overlapping galaxy redshift survey, showing it close to unity rather than the factor of several the model predicts, would break the forecast. A realistic foreground-removal study showing that the $ell < 60$ modes are lost would do the same.","tokens_in":11042,"feed_emoji":"📡","tokens_out":2435,"duration_ms":56016,"temperature":0.7,"pith_summary":"The paper asks whether a survey that maps the collective 21 cm emission of neutral hydrogen, without resolving individual galaxies, can measure the scale-dependent signature that primordial non-Gaussianity imprints on the clustering of matter. The authors argue that the answer is yes, in principle: a purpose-built low-frequency intensity mapping experiment covering $z$ between roughly 0.5 and 4.5 could reach an uncertainty on the non-Gaussianity amplitude $f_{NL}$ of order 1. The prize is a constraint on a parameter that separates canonical single-field inflation, which predicts $|f_{NL}|$ of order $10^{-2}$, from models that generate values of order 1 through post-inflationary evolution. The argument rests on the fact that intensity mapping reaches very large spatial scales at high redshift, where the non-Gaussian signal is strongest and where conventional galaxy surveys are slow and cosmic-variance limited.","feed_headline":"Hydrogen maps could pin inflation's non-Gaussianity to ±1","feed_subtitle":"A low-frequency 21 cm survey spanning redshifts 1 to 5 could beat Planck's f_NL bound by an order of magnitude.","key_machinery":"The load-bearing mechanism is the scale-dependent bias induced by primordial non-Gaussianity: a tracer's bias acquires a correction $Delta b_X(z,k) = 3 (b^G_X(z) - 1) Omega_m H_0^2 delta_c / [c^2 k^2 T(k) D_+(z)] f_{NL}$, which grows toward small $k$ and vanishes for tracers with bias equal to one. The authors model the neutral hydrogen bias by integrating a Sheth-Tormen halo mass function with hydrogen mass proportional to halo mass, bounded by a circular-velocity cutoff, and normalising to an external measurement of $Omega_{HI} b_{HI}$. The forecast itself is a Fisher matrix built from tomographic 21 cm angular power spectra computed with the CAMB-sources code, including redshift-space distortions and a top-hat frequency window, with noise models for both dish arrays and interferometers.","core_discovery":"The central claim is a forecast, not a detection. The authors compute the angular power spectrum of 21 cm intensity fluctuations in redshift slices, add the standard scale-dependent correction to the bias that a non-zero $f_{NL}$ induces, and run a Fisher matrix over the tomographic spectrum to estimate the marginal error on $f_{NL}$. For a cosmic-variance-limited survey, the error scales as $[b^G_X T(k_alpha) D_+(z) c^2]^2 / ([6(b^G_X - 1) Omega_m H_0^2 delta_c]^2 sum_alpha N_alpha k_alpha^{-4})$, so two things drive the result: the number of modes, which grows with survey volume and depth, and how far the tracer's bias sits from unity. Neutral hydrogen is underbiased at low redshift and strongly biased at early times, and the authors show that a deep, wide survey reaching frequencies below about 400 MHz is what makes $sigma_{f_{NL}} sim 1$ achievable. They then translate that requirement into concrete instrument specifications, noting that a dish array with roughly 10,000 small elements near 400 MHz, or an interferometer with a field of view near 1000 square degrees, would be needed; existing and planned instruments such as MeerKAT, APERTIF, ASKAP and GMRT fall short on frequency coverage, sensitivity or field of view.","pith_inferences":["The forecast's sensitivity to the bias evolution suggests a cheap intermediate test: cross-correlating 21 cm intensity maps with an overlapping galaxy survey at $z sim 1$-$2$ would measure the effective hydrogen bias directly, and that measurement could be fed back into the Fisher analysis before any dedicated instrument is built.","The authors treat foreground removal as tractable because Galactic synchrotron emission varies slowly in frequency, but the largest angular scales are exactly where the $f_{NL}$ signal lives; a dedicated study of how much of the $ell < 60$ range survives realistic cleaning would be the natural next step and could shift the quoted error substantially.","If the hydrogen bias at $z sim 3$-$4$ turned out to be closer to one than the Sheth-Tormen model predicts, the same survey would still constrain $f_{NL}$, but the required observing time would grow faster than the survey area, which changes which instrument concept is cheapest rather than whether the measurement is possible.","The scale-dependent bias signature is shared by any tracer of the dark matter, so the intensity mapping case is best read as an argument that the very largest scales are worth instrumenting at all, independent of which tracer is eventually used."],"forward_implications":["A measurement of $sigma_{f_{NL}} sim 1$ would discriminate between canonical single-field inflation and models producing order-unity non-Gaussianity through post-inflationary evolution, a distinction the cosmic microwave background alone cannot make at that precision.","The same survey would deliver a three-dimensional map of the matter distribution on ultralarge scales out to $z sim 4$, useful for testing general-relativistic and gauge effects that become non-negligible on those scales.","The forecast sets a concrete design target for a dedicated instrument: a compact array working near 400 MHz with a wide field of view and a large bandwidth, rather than the high angular resolution that current interferometer designs prioritise.","Because the constraint improves as the tracer bias moves away from unity, the result argues for spanning a broad redshift range rather than optimising a single deep slice."],"supporting_citations":[{"why":"Supplies the scale-dependent bias correction induced by primordial non-Gaussianity, the effect the forecast is built to detect.","marker":"[22]"},{"why":"Independent derivation of the same scale-dependent bias signature, underpinning the signal model.","marker":"[23]"},{"why":"Provides the CAMB-sources code and the angular power spectrum formalism used to compute the tomographic 21 cm spectra.","marker":"[10]"},{"why":"Supplies the Sheth-Tormen mass function and halo bias used to model the neutral hydrogen bias.","marker":"[31]"},{"why":"Provides the measurement of $Omega_{HI} b_{HI}$ used to normalise the hydrogen abundance and bias model.","marker":"[8]"},{"why":"Establishes the Fisher matrix formalism applied to the tomographic angular power spectra.","marker":"[34]"},{"why":"Gives the Planck constraint on $f_{NL}$ that the forecasted intensity mapping bound is compared against.","marker":"[21]"},{"why":"Introduces the intensity mapping approach of mapping unresolved 21 cm emission that the survey concept relies on.","marker":"[4]"},{"why":"Supports the claim that foreground cleaning affects only scales set by the bandwidth used for removal, which the analysis depends on.","marker":"[35]"}],"fun_headline_variants":["21 cm intensity map forecast: f_NL error near 1","Proposed hydrogen survey could bound f_NL to about 1","Deep 21 cm scan may reach f_NL uncertainty of 1","Wide neutral hydrogen map could tighten f_NL to ±1"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The forecast assumes a specific, unmeasured redshift evolution of the neutral hydrogen bias, derived from a halo mass function with hydrogen mass proportional to halo mass; if the real bias stays closer to one at high redshift, the projected error on $f_{NL}$ degrades well beyond the quoted value of about 1.","fun_headline_variants_meta":{"raw":{"variants":["21 cm intensity map forecast: f_NL error near 1","Proposed hydrogen survey could bound f_NL to about 1","Deep 21 cm scan may reach f_NL uncertainty of 1","Wide neutral hydrogen map could tighten f_NL to ±1"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":2028,"prompt_tokens":949,"completion_tokens":1079,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":1014}},"tokens_in":949,"tokens_out":1079,"duration_ms":19697,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":1014,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-12T22:10:40.388817+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"pipeline_version":"pith-pipeline@v1.5.0-archive-selfhost","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"A direct measurement of the neutral hydrogen bias at $z$ between 2 and 4, for instance from cross-correlating an intensity map with an overlapping galaxy redshift survey, showing it close to unity rather than the factor of several the model predicts, would break the forecast. A realistic foreground-removal study showing that the $ell < 60$ modes are lost would do the same.","supporting_citations":[{"cited_title":"The imprints of primordial non-gaussianities on large-scale structure: scale dependent bias and abundance of virialized objects","cited_arxiv_id":"0710.4560","evidence_quote":"Supplies the scale-dependent bias correction induced by primordial non-Gaussianity, the effect the forecast is built to detect."},{"cited_title":"The effect of primordial non-Gaussianity on halo bias","cited_arxiv_id":"0801.4826","evidence_quote":"Independent derivation of the same scale-dependent bias signature, underpinning the signal model."},{"cited_title":"The linear power spectrum of observed source number counts","cited_arxiv_id":"1105.5292","evidence_quote":"Provides the CAMB-sources code and the angular power spectrum formalism used to compute the tomographic 21 cm spectra."},{"cited_title":"Ellipsoidal collapse and an improved model for the number and spatial distribution of dark matter haloes","cited_arxiv_id":"astro-ph/9907024","evidence_quote":"Supplies the Sheth-Tormen mass function and halo bias used to model the neutral hydrogen bias."},{"cited_title":"Determination of z~0.8 neutral hydrogen fluctuations using the 21 cm intensity mapping auto-correlation","cited_arxiv_id":"1304.3712","evidence_quote":"Provides the measurement of $Omega_{HI} b_{HI}$ used to normalise the hydrogen abundance and bias model."},{"cited_title":"Karhunen-Loeve eigenvalue problems in cosmology: how should we tackle large data sets?","cited_arxiv_id":"astro-ph/9603021","evidence_quote":"Establishes the Fisher matrix formalism applied to the tomographic angular power spectra."},{"cited_title":"Planck 2013 Results. XXIV. Constraints on primordial non-Gaussianity","cited_arxiv_id":"1303.5084","evidence_quote":"Gives the Planck constraint on $f_{NL}$ that the forecasted intensity mapping bound is compared against."},{"cited_title":"Baryon Acoustic Oscillation Intensity Mapping as a Test of Dark Energy","cited_arxiv_id":"0709.3672","evidence_quote":"Introduces the intensity mapping approach of mapping unresolved 21 cm emission that the survey concept relies on."},{"cited_title":"Foreground Removal using FastICA: A Showcase of LOFAR-EoR","cited_arxiv_id":"1201.2190","evidence_quote":"Supports the claim that foreground cleaning affects only scales set by the bandwidth used for removal, which the analysis depends on."}],"lane_accuracy":{"status":"not yet measured","lane":{"model":"selfhosted/deepseek-v4.1-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}