© 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. 9967
    Prev
    Next
    INDEX
    Explanations

    sections related to webpage navigation or content management

    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
    111
    +0.37
    2.3%
    156
    +0.36
    2.2%
    307
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    298
    +0.37
    0.03
    156
    +0.36
    0.02
    473
    +0.10
    0.02
    Negative Logits
    ĥ½
    -4.89
    <|outofrange|>
    -4.68
    <|outofrange|>
    -4.68
    <|outofrange|>
    -4.68
    <|padding|>
    -4.68
    ↵
    -4.68
    <|outofrange|>
    -4.68
    -4.68
    <|outofrange|>
    -4.68
     ↵
    -4.68
    POSITIVE LOGITS
    quare
    1.62
    esome
    1.59
    ://
    1.47
    ety
    1.46
     Books
    1.43
    ples
    1.42
    "}](#
    1.40
    edly
    1.39
    enden
    1.39
    eded
    1.35
    Activations Density 0.046%

    No Known Activations