© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 10373
    Prev
    Next
    INDEX
    Explanations

    quantitative measurements and concentrations related to scientific experiments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/4-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.4.hook_resid_post
    Hook Layer
    4
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    369
    +0.13
    0.7%
    186
    +0.12
    0.7%
    94
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    458
    +0.13
    0.03
    470
    +0.12
    0.05
    477
    +0.11
    0.04
    Negative Logits
     eligibility
    -1.67
    ariat
    -1.59
    arial
    -1.51
    aries
    -1.47
     Surgery
    -1.44
    :**
    -1.44
    arias
    -1.41
    tica
    -1.40
    idas
    -1.39
    artment
    -1.39
    POSITIVE LOGITS
    ħ
    3.25
    ³
    2.52
    ¬
    2.44
    °
    2.41
    Ń
    2.39
    ª
    2.38
    %
    2.34
     ppm
    2.33
    kDa
    2.32
    Ĵ
    2.31
    Activations Density 0.154%

    No Known Activations