© 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. 29576
    Prev
    Next
    INDEX
    Explanations

    terms related to the extraction and purification of substances

    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
    +0.21
    1.2%
    89
    +0.12
    0.7%
    203
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.21
    0.08
    156
    +0.12
    0.05
    271
    +0.12
    0.03
    Negative Logits
    ĥ½
    -2.29
    ¾
    -1.93
    §
    -1.88
    Ī
    -1.85
    Ĥ¬
    -1.85
    Ĩ
    -1.84
    ĸ´
    -1.71
     rede
    -1.70
    ¼
    -1.68
    Į
    -1.66
    POSITIVE LOGITS
     ensures
    1.47
    TE
    1.37
    ooth
    1.35
     guarantees
    1.35
    Tip
    1.33
     allows
    1.31
     costs
    1.31
     involves
    1.30
     discloses
    1.30
    veolar
    1.30
    Activations Density 0.255%

    No Known Activations