© 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. 1-RES-SM
    11. 24705
    Prev
    Next
    INDEX
    Explanations

    variations of the word "run."

    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
    184
    +0.15
    0.8%
    162
    +0.14
    0.8%
    376
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    162
    +0.15
    0.05
    184
    +0.14
    0.05
    342
    +0.13
    0.05
    Negative Logits
     willing
    -1.84
    ł
    -1.67
    ľ
    -1.50
     who
    -1.47
    º
    -1.45
    Ļ
    -1.44
    ŀ
    -1.39
    ¥
    -1.37
     abuse
    -1.35
     senses
    -1.35
    POSITIVE LOGITS
    away
    2.38
    aways
    2.05
    offs
    1.95
    swick
    1.84
    emann
    1.78
    guns
    1.69
    heimer
    1.65
     af
    1.65
    abouts
    1.60
    yon
    1.57
    Activations Density 0.095%

    No Known Activations