© 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. 10519
    Prev
    Next
    INDEX
    Explanations

    code snippets or programming-related syntax elements

    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
    76
    +0.13
    0.7%
    194
    +0.13
    0.7%
    441
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    212
    +0.13
    0.04
    453
    +0.13
    0.03
    441
    +0.13
    0.00
    Negative Logits
    Ļª
    -2.67
    ĻĤ
    -2.20
    Ĭ
    -2.16
    Ļ
    -2.15
    °
    -2.15
    ´
    -2.07
    ĥ½
    -1.95
    »¿
    -1.93
    ħ
    -1.89
    ³
    -1.88
    POSITIVE LOGITS
     add
    3.56
     Add
    3.20
    Add
    2.99
    add
    2.97
     subtract
    2.83
    added
    2.74
     ADD
    2.69
     added
    2.67
    Added
    2.61
    Adding
    2.51
    Activations Density 0.422%

    No Known Activations