© 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. 27099
    Prev
    Next
    INDEX
    Explanations

    recurring phrases that emphasize possession or relationships

    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
    23
    +0.21
    1.2%
    419
    +0.14
    0.8%
    478
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    419
    +0.21
    0.04
    161
    +0.14
    0.05
    295
    +0.14
    0.04
    Negative Logits
    ī
    -3.05
    Ĩ
    -2.69
    ¡
    -2.50
    ģ
    -2.47
    Ĭ
    -2.44
    Ģ
    -2.43
    Ī
    -2.38
    İ
    -2.37
    ĭ
    -2.37
    ĩ
    -2.33
    POSITIVE LOGITS
     matter
    1.80
    opy
    1.71
     Happ
    1.57
    INCLUDING
    1.50
     life
    1.48
     relation
    1.47
    ometh
    1.47
     individual
    1.46
     physical
    1.46
     young
    1.45
    Activations Density 0.216%

    No Known Activations