© 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 EvalsExportsAPI 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. 10441
    Prev
    Next
    INDEX
    Explanations

    references to siblings and familial relationships

    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
    94
    +0.14
    0.8%
    156
    +0.10
    0.6%
    410
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    94
    +0.14
    0.03
    56
    +0.10
    -0.00
    413
    +0.10
    0.01
    Negative Logits
    ureus
    -2.01
    PROM
    -1.61
    urin
    -1.55
    ibel
    -1.54
    urized
    -1.47
    legraph
    -1.46
    uria
    -1.44
    CTOR
    -1.43
    **](#
    -1.41
    illery
    -1.40
    POSITIVE LOGITS
    hood
    2.72
    stown
    2.05
    ships
    1.93
     footsteps
    1.75
     rivalry
    1.71
    bery
    1.69
     ranks
    1.67
    liness
    1.67
    ially
    1.57
     sexes
    1.56
    Activations Density 0.331%

    No Known Activations