© 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. 30853
    Prev
    Next
    INDEX
    Explanations

    terms related to statistical distribution

    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
    217
    +0.14
    0.8%
    309
    +0.12
    0.7%
    471
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    29
    +0.14
    0.01
    471
    +0.12
    0.01
    361
    +0.10
    0.01
    Negative Logits
    ĭ
    -3.67
    µ
    -3.54
    ¹
    -3.54
    Į
    -3.53
    §
    -3.48
    º
    -3.46
    ĥ½
    -3.41
    ¤
    -3.40
    «
    -3.36
    ĸ
    -3.26
    POSITIVE LOGITS
     consist
    1.81
    ination
    1.74
    inator
    1.71
    inates
    1.70
     purposes
    1.62
    al
    1.60
    ality
    1.60
     among
    1.57
     consists
    1.57
    chart
    1.57
    Activations Density 0.067%

    No Known Activations