© 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. 2-RES-SM
    11. 10635
    Prev
    Next
    INDEX
    Explanations

    references to marital relationships, particularly the term "wife."

    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
    410
    +0.13
    0.7%
    328
    +0.12
    0.7%
    163
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    163
    +0.13
    0.03
    56
    +0.12
    0.01
    410
    +0.12
    0.05
    Negative Logits
    ĵ
    -2.84
    ĸ
    -2.24
    ¯
    -1.86
    Ĥ¬
    -1.81
    ¾
    -1.79
    amers
    -1.79
    ¬
    -1.77
    ³
    -1.76
    ł
    -1.74
    ½
    -1.74
    POSITIVE LOGITS
     herself
    2.64
     Elizabeth
    1.99
    hood
    1.87
     Helen
    1.81
     Nina
    1.75
    liness
    1.75
     quarters
    1.75
     Esther
    1.71
     actress
    1.68
    dess
    1.68
    Activations Density 0.305%

    No Known Activations