© 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. 27035
    Prev
    Next
    INDEX
    Explanations

    instances of the word "Mon" in various contexts

    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
    369
    +0.17
    0.9%
    112
    +0.13
    0.8%
    349
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    112
    +0.17
    0.03
    477
    +0.13
    0.03
    369
    +0.13
    0.01
    Negative Logits
    lass
    -1.67
    appropriate
    -1.63
    lessly
    -1.59
    ESULT
    -1.57
    arynge
    -1.55
    rapeut
    -1.53
     poor
    -1.51
     authenticity
    -1.50
     diagnosis
    -1.49
    self
    -1.49
    POSITIVE LOGITS
    ths
    2.04
    ettes
    1.98
    nes
    1.93
    estones
    1.79
    zes
    1.77
    agle
    1.65
    ¿
    1.64
    astery
    1.63
    estone
    1.62
    ¾
    1.60
    Activations Density 0.105%

    No Known Activations