© 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. 30925
    Prev
    Next
    INDEX
    Explanations

    the concept of sharing

    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
    29
    +0.13
    0.7%
    108
    +0.12
    0.7%
    239
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    29
    +0.13
    0.02
    239
    +0.12
    0.03
    108
    +0.11
    0.03
    Negative Logits
    ľĵ
    -2.50
    ĨĴ
    -2.35
    ij
    -2.31
    Ĩ
    -2.26
    ½
    -2.24
    Ļª
    -2.24
    ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ
    -2.22
    ¹
    -2.22
    ¼
    -2.22
      ↵
    -2.21
    POSITIVE LOGITS
    holder
    1.99
    holding
    1.76
    cro
    1.71
    able
    1.67
    ership
    1.66
     bases
    1.63
     ownership
    1.59
    creen
    1.59
    quot
    1.57
     bathrooms
    1.51
    Activations Density 0.178%

    No Known Activations