© 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. 1-RES-SM
    11. 28014
    Prev
    Next
    INDEX
    Explanations

    references to legal citations and case law

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    69
    +0.18
    1.0%
    351
    +0.11
    0.6%
    317
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    69
    +0.18
    0.08
    81
    +0.11
    0.06
    151
    +0.11
    0.05
    Negative Logits
    ĻĤ
    -2.75
    ©
    -2.06
    ĺ
    -1.96
    ĸ´
    -1.82
    unnumbered
    -1.69
    Īĺ
    -1.63
    ķ
    -1.58
    ¿½
    -1.57
    ¤
    -1.50
    ľ
    -1.46
    POSITIVE LOGITS
    ricular
    2.00
    naire
    1.99
    usiness
    1.89
    ional
    1.66
    asone
    1.61
    cil
    1.60
    icle
    1.54
    enue
    1.52
    velopment
    1.52
    arian
    1.52
    Activations Density 0.325%

    No Known Activations