© 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. 10538
    Prev
    Next
    INDEX
    Explanations

    sections of code documentation, particularly headers and comments

    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
    156
    +0.19
    1.1%
    478
    +0.13
    0.7%
    304
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    192
    +0.19
    0.02
    304
    +0.13
    0.02
    347
    +0.13
    0.02
    Negative Logits
    territory
    -1.71
    pressing
    -1.55
    ressor
    -1.52
    pled
    -1.51
    otine
    -1.46
    spot
    -1.42
    blast
    -1.38
    atinib
    -1.37
    drive
    -1.37
     territory
    -1.35
    POSITIVE LOGITS
    ¿½
    1.94
     nm
    1.67
     forth
    1.62
     eth
    1.55
    :--
    1.55
    openh
    1.43
     anybody
    1.43
     Copyright
    1.42
     anecd
    1.41
     cref
    1.38
    Activations Density 0.133%

    No Known Activations