© 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. 30915
    Prev
    Next
    INDEX
    Explanations

    technical specifications and details related to products or systems

    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
    391
    +0.17
    1.0%
    350
    +0.16
    1.0%
    386
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    350
    +0.17
    0.03
    386
    +0.16
    0.01
    391
    +0.14
    0.02
    Negative Logits
    ´
    -3.39
    ĩ
    -2.96
    ľĵ
    -2.90
    °
    -2.79
    ¸
    -2.79
    ĥ½
    -2.77
    §
    -2.75
    µ
    -2.67
    ĭ
    -2.66
    ½
    -2.64
    POSITIVE LOGITS
    ials
    2.15
    ifically
    2.07
    iousness
    1.92
    ifiers
    1.92
    ification
    1.86
    ulations
    1.85
    ifi
    1.84
    iation
    1.83
    ifications
    1.83
    ified
    1.79
    Activations Density 0.187%

    No Known Activations