© 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. 31935
    Prev
    Next
    INDEX
    Explanations

    instances of the word "software" in various contexts related to licensing and usage rights

    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
    364
    +0.11
    0.6%
    17
    +0.11
    0.6%
    453
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    453
    +0.11
    0.02
    364
    +0.11
    0.02
    107
    +0.11
    0.01
    Negative Logits
    ¯
    -1.91
    ®
    -1.88
    ĻĤ
    -1.82
    most
    -1.78
    ĨĴ
    -1.72
    ı
    -1.69
     wonder
    -1.65
    ľĵ
    -1.65
    Ļª
    -1.64
    Ĺ
    -1.58
    POSITIVE LOGITS
     itself
    1.72
     Amendment
    1.64
     Agreement
    1.60
    ician
    1.51
     Engineering
    1.50
     Initiative
    1.48
     bases
    1.47
    bourg
    1.46
     Foundation
    1.45
     Collaboration
    1.43
    Activations Density 0.037%

    No Known Activations