© 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. 30752
    Prev
    Next
    INDEX
    Explanations

    repeated instances of the definite article "the."

    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
    478
    +0.21
    1.2%
    369
    +0.18
    1.1%
    98
    +0.18
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    10
    +0.21
    0.06
    467
    +0.18
    0.03
    444
    +0.18
    0.02
    Negative Logits
    "}](#
    -2.02
    CUIT
    -1.73
    ENTIAL
    -1.71
    ionale
    -1.66
     centrifugation
    -1.60
    LECT
    -1.59
    )}^{(
    -1.54
    arroll
    -1.53
    uchi
    -1.52
    PROP
    -1.50
    POSITIVE LOGITS
    ĨĴ
    2.77
     latter
    2.24
    ĵ
    2.20
    ↵
    2.14
    ↵↵
    2.14
    ↵↵
    2.14
    č↵č↵
    2.14
    ↵  Âł
    2.14
     ↵
    2.14
    ↵
    2.14
    Activations Density 0.321%

    No Known Activations