© 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. 1-RES-SM
    11. 29182
    Prev
    Next
    INDEX
    Explanations

    brackets and code closure elements in programming or markup languages

    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
    173
    +0.12
    0.7%
    478
    +0.10
    0.6%
    225
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.12
    0.06
    9
    +0.10
    0.04
    225
    +0.10
    0.05
    Negative Logits
    unreadable
    -1.69
    than
    -1.68
     perspectives
    -1.56
     prizes
    -1.55
    kinase
    -1.51
    guards
    -1.47
    YPT
    -1.47
     modalities
    -1.44
    walks
    -1.44
     workload
    -1.43
    POSITIVE LOGITS
    Ĩ
    1.85
    ķ
    1.68
    ¿½
    1.67
     fact
    1.66
    «
    1.57
    ĻĤ
    1.56
    ľĵ
    1.53
    ¦
    1.53
    ĥ½
    1.52
    ģ
    1.50
    Activations Density 0.183%

    No Known Activations