© 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. 24372
    Prev
    Next
    INDEX
    Explanations

    references to Microsoft Excel and its features

    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
    +0.20
    1.2%
    490
    +0.13
    0.7%
    139
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    321
    +0.20
    0.01
    392
    +0.13
    0.01
    225
    +0.10
    0.01
    Negative Logits
     fixed
    -1.69
     autism
    -1.65
     transgender
    -1.59
     currents
    -1.54
     pedestrians
    -1.53
     adopted
    -1.45
     superior
    -1.45
     abusive
    -1.44
    athered
    -1.43
     teens
    -1.43
    POSITIVE LOGITS
    ière
    1.91
    erals
    1.90
    ente
    1.88
    »¿
    1.76
    writer
    1.74
    etable
    1.72
    nement
    1.72
    processor
    1.71
    ery
    1.67
    ious
    1.67
    Activations Density 0.087%

    No Known Activations