© 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. 26503
    Prev
    Next
    INDEX
    Explanations

    instances of the word "had"

    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
    23
    +0.23
    1.3%
    478
    +0.16
    0.9%
    376
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    304
    +0.23
    0.07
    132
    +0.16
    0.07
    356
    +0.12
    0.06
    Negative Logits
    ł
    -3.64
    Ļª
    -3.47
    »¿
    -3.38
    ½
    -3.20
    £
    -3.13
    ĭ
    -3.09
    Ģ
    -2.97
    Ŀ
    -2.95
    Ń
    -2.94
    µ
    -2.92
    POSITIVE LOGITS
    dock
    2.22
    oop
    2.14
    rien
    1.76
    rian
    1.62
    asms
    1.60
     recently
    1.59
    itor
    1.58
     previously
    1.55
     Parenthood
    1.55
    boys
    1.53
    Activations Density 0.188%

    No Known Activations