© 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. 4-RES-SM
    11. 13317
    Prev
    Next
    INDEX
    Explanations

    references to software licensing and permissions

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    78
    +0.15
    0.9%
    57
    +0.13
    0.7%
    166
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    57
    +0.15
    0.03
    162
    +0.13
    0.01
    391
    +0.13
    0.02
    Negative Logits
    ¾
    -1.55
    ulum
    -1.51
     because
    -1.50
    edin
    -1.49
    leneck
    -1.47
     axis
    -1.45
    hesis
    -1.44
     governor
    -1.42
    iscus
    -1.40
    ului
    -1.40
    POSITIVE LOGITS
    ublic
    1.69
    áĢ¯
    1.65
    áĢIJ
    1.52
    র
    1.44
     Materials
    1.44
    perty
    1.43
    aceous
    1.42
     Software
    1.41
     conduct
    1.40
    à¦
    1.40
    Activations Density 0.083%

    No Known Activations