© 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 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. 27472
    Prev
    Next
    INDEX
    Explanations

    instances of file-related terminology

    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
    91
    +0.17
    1.0%
    93
    +0.14
    0.8%
    100
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    91
    +0.17
    0.04
    93
    +0.14
    0.03
    181
    +0.13
    0.02
    Negative Logits
    ·
    -1.71
    ¯
    -1.71
    ĥ½
    -1.70
    ³
    -1.69
    ¤
    -1.64
    µ
    -1.57
    ķ
    -1.57
    Ń
    -1.53
    Ĺ
    -1.53
    iph
    -1.53
    POSITIVE LOGITS
    zilla
    1.89
     cabinets
    1.85
    =/
    1.65
    balls
    1.64
    Exists
    1.64
    's
    1.63
    roit
    1.61
    ÃŃs
    1.60
    sworth
    1.58
     contains
    1.58
    Activations Density 0.071%

    No Known Activations