© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 11074
    Prev
    Next
    INDEX
    Explanations

    instances of severe health issues or conditions

    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
    326
    +0.16
    0.9%
    71
    +0.13
    0.7%
    240
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.16
    0.02
    418
    +0.13
    0.00
    493
    +0.12
    0.03
    Negative Logits
     fare
    -1.66
    icks
    -1.49
    strand
    -1.46
    place
    -1.43
    ).)
    -1.42
     brackets
    -1.42
     dice
    -1.41
    ction
    -1.37
    ?_
    -1.34
    ware
    -1.33
    POSITIVE LOGITS
    »
    1.84
     pharmacological
    1.79
    arth
    1.79
    Ī
    1.71
    Į
    1.67
    Population
    1.56
    »¿
    1.52
    ĵ
    1.50
    Ģ
    1.49
    urable
    1.49
    Activations Density 0.129%

    No Known Activations