© 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. 26442
    Prev
    Next
    INDEX
    Explanations

    mentions of the city Houston and its surrounding areas

    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
    156
    +0.19
    1.1%
    182
    +0.17
    1.0%
    71
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    182
    +0.19
    0.04
    71
    +0.17
    0.04
    321
    +0.14
    0.03
    Negative Logits
    ĨĴ
    -2.20
     ours
    -1.88
    ¦
    -1.86
    ĩ
    -1.86
     yours
    -1.81
    ı
    -1.81
    Ĩ
    -1.73
    Īĺ
    -1.72
    Ħ
    -1.71
    ĺ
    -1.68
    POSITIVE LOGITS
    iana
    1.80
     Texans
    1.78
     Rapids
    1.71
    burn
    1.71
    pler
    1.70
    iej
    1.61
    iang
    1.51
    shire
    1.50
    ian
    1.49
    icidal
    1.49
    Activations Density 0.132%

    No Known Activations