© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 29949
    Prev
    Next
    INDEX
    Explanations

    PHP code snippets

    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
    410
    +0.14
    0.8%
    315
    +0.14
    0.8%
    478
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    315
    +0.14
    0.01
    29
    +0.14
    0.01
    228
    +0.13
    0.01
    Negative Logits
    oint
    -1.92
    icense
    -1.82
    oides
    -1.70
    onal
    -1.66
    )</
    -1.64
    ()</
    -1.64
    arent
    -1.61
    uke
    -1.59
    edia
    -1.56
    ief
    -1.56
    POSITIVE LOGITS
    ¦
    2.48
    Ľ
    2.29
    ĭ
    2.08
    ¨
    1.97
    ģ
    1.95
    IJ
    1.90
    ĸ
    1.82
    ©
    1.79
    ł
    1.71
    ¸
    1.70
    Activations Density 0.071%

    No Known Activations