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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 29247
    Prev
    Next
    INDEX
    Explanations

    URLs or hyperlinks in the text

    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
    478
    +0.14
    0.8%
    315
    +0.13
    0.7%
    197
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    315
    +0.14
    0.04
    462
    +0.13
    0.03
    458
    +0.11
    0.04
    Negative Logits
    manship
    -1.80
    fected
    -1.55
    zzles
    -1.54
     transmitted
    -1.50
     function
    -1.49
     well
    -1.49
    borne
    -1.46
     malfunction
    -1.46
    ?”
    -1.44
     hereto
    -1.43
    POSITIVE LOGITS
    ĥ½
    3.17
    ½
    2.95
    Ŀ
    2.95
    ¿
    2.91
    ¾
    2.88
    ¤
    2.83
    İ
    2.80
    ¦
    2.79
    ĨĴ
    2.77
    ¥
    2.76
    Activations Density 0.137%

    No Known Activations