Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 28882
    Prev
    Next
    INDEX
    Explanations

    terms related to pediatric and obstetric care

    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
    156
    +0.28
    1.6%
    87
    +0.15
    0.9%
    34
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    34
    +0.28
    0.01
    87
    +0.15
    0.01
    18
    +0.13
    0.01
    Negative Logits
    uben
    -1.65
     direct
    -1.53
     redeem
    -1.48
     true
    -1.32
    unto
    -1.32
    ayed
    -1.31
    akes
    -1.31
     Boolean
    -1.31
    iber
    -1.31
     conflicting
    -1.30
    POSITIVE LOGITS
    ĸ
    2.89
    ł
    2.80
    §
    2.79
    £
    2.70
    º
    2.59
    Ĩ
    2.56
    ·¸
    2.54
    Ł
    2.52
    ¨
    2.52
    Ĥ
    2.51
    Activations Density 0.065%

    No Known Activations