© 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. 2-RES-SM
    11. 26687
    Prev
    Next
    INDEX
    Explanations

    specific product features and program details related to subscriptions and loans

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    354
    +0.14
    0.7%
    78
    +0.13
    0.7%
    210
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    354
    +0.14
    0.09
    78
    +0.13
    0.05
    174
    +0.11
    0.10
    Negative Logits
     LLP
    -1.93
    idone
    -1.60
    otec
    -1.54
    glich
    -1.52
    ãģ¾ãģĽãĤĵ
    -1.44
    held
    -1.42
    perform
    -1.41
     Rica
    -1.39
    sein
    -1.39
     rid
    -1.39
    POSITIVE LOGITS
     consists
    1.91
     lasts
    1.82
     itself
    1.81
     contains
    1.77
     consisted
    1.75
     divides
    1.63
     lasted
    1.62
     goes
    1.61
     continues
    1.54
     appears
    1.54
    Activations Density 1.942%

    No Known Activations