© 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. 13841
    Prev
    Next
    INDEX
    Explanations

    references to military honors and awards, particularly those related to health and safety issues

    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
    78
    +0.15
    0.8%
    452
    +0.12
    0.7%
    202
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    22
    +0.15
    0.06
    202
    +0.12
    -0.00
    137
    +0.12
    0.11
    Negative Logits
    hip
    -1.95
    pancy
    -1.86
    fecture
    -1.69
    culation
    -1.60
    urity
    -1.59
    iciency
    -1.59
    bidity
    -1.58
    ción
    -1.57
    gmail
    -1.56
    áĢº
    -1.55
    POSITIVE LOGITS
    ħ
    4.19
    Ľ
    4.06
    ĥ
    3.97
    ł
    3.88
    ľĵ
    3.72
    ij
    3.71
    Ļ
    3.71
    ¨
    3.71
    Ł
    3.68
    ı
    3.66
    Activations Density 1.720%

    No Known Activations