© 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. 1-RES-SM
    11. 27066
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "constructor"

    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%
    488
    +0.15
    0.8%
    365
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    73
    +0.19
    0.02
    467
    +0.15
    0.02
    488
    +0.13
    0.01
    Negative Logits
    ĥ½
    -4.27
    ĥ
    -4.11
    ¦
    -4.06
    º
    -4.04
    ·
    -4.01
    į
    -4.00
    -3.92
    <|outofrange|>
    -3.92
    -3.92
    -3.92
    POSITIVE LOGITS
    als
    1.67
     underway
    1.59
    ality
    1.55
    isions
    1.54
    amps
    1.54
    fires
    1.49
     rule
    1.47
    idable
    1.47
    AtIndex
    1.46
    amped
    1.46
    Activations Density 0.099%

    No Known Activations