© 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. 12504
    Prev
    Next
    INDEX
    Explanations

    numerical data and patterns

    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
    69
    +0.14
    0.8%
    203
    +0.13
    0.7%
    274
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    477
    +0.14
    0.07
    69
    +0.13
    0.06
    136
    +0.13
    -0.01
    Negative Logits
    eting
    -2.03
    holders
    -1.75
    inations
    -1.75
    holder
    -1.72
    ually
    -1.59
    hold
    -1.59
    unt
    -1.51
    ning
    -1.49
    alities
    -1.35
    mouth
    -1.35
    POSITIVE LOGITS
    Ļ
    1.86
    ´
    1.83
    °
    1.81
    Į
    1.75
    ļ
    1.74
    ±
    1.72
    ©
    1.71
    ·
    1.70
    ŀ
    1.67
    Ł
    1.62
    Activations Density 0.322%

    No Known Activations