© 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. 30680
    Prev
    Next
    INDEX
    Explanations

    references to oxygen and hydrogen 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
    156
    +0.18
    1.0%
    320
    +0.14
    0.8%
    274
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    275
    +0.18
    0.01
    446
    +0.14
    0.01
    320
    +0.13
    0.01
    Negative Logits
     href
    -1.84
     ever
    -1.83
     gay
    -1.81
    ère
    -1.75
    lı
    -1.72
     married
    -1.71
    yahoo
    -1.70
    yll
    -1.66
    érie
    -1.65
     sister
    -1.65
    POSITIVE LOGITS
    ħ
    2.71
    İ
    2.65
    Ĵ
    2.55
    ģ
    2.40
    ľ
    2.31
    ¨
    2.30
    Ħ
    2.30
    ¢
    2.28
    ĵ
    2.18
    ij
    2.17
    Activations Density 0.055%

    No Known Activations