© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 26773
    Prev
    Next
    INDEX
    Explanations

    terms and phrases related to laser technology and treatments

    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
    156
    +0.24
    1.4%
    148
    +0.14
    0.8%
    443
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    443
    +0.24
    0.01
    148
    +0.14
    0.00
    304
    +0.12
    0.01
    Negative Logits
    ylase
    -1.60
     grammat
    -1.59
    emic
    -1.58
    lla
    -1.54
     stride
    -1.50
    consist
    -1.47
    ordinate
    -1.45
    ção
    -1.44
     Jewish
    -1.43
     entirety
    -1.43
    POSITIVE LOGITS
    spot
    1.65
    KERN
    1.64
    film
    1.61
     directed
    1.59
    LEY
    1.54
    craft
    1.51
    olta
    1.48
    EXPORT
    1.48
    horn
    1.48
    ue
    1.48
    Activations Density 0.080%

    No Known Activations