© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 11410
    Prev
    Next
    INDEX
    Explanations

    geographical locations and street names

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    25
    +0.20
    1.3%
    9
    +0.17
    1.1%
    35
    +0.16
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    9
    +0.20
    0.13
    25
    +0.17
    0.22
    325
    +0.16
    0.18
    Negative Logits
    orld
    -1.39
    izumab
    -1.28
    000001
    -1.27
    interface
    -1.15
    FTWARE
    -1.12
     approached
    -1.12
     WARRANTY
    -1.11
     regards
    -1.11
     than
    -1.10
    isco
    -1.09
    POSITIVE LOGITS
    č↵
    1.61
     ↵
    1.61
    1.61
     ↵
    1.61
    1.61
    č↵
    1.61
    ↵↵
    1.61
    ↵
    1.61
    ↵↵
    1.61
    1.61
    Activations Density 5.705%

    No Known Activations