© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 12165
    Prev
    Next
    INDEX
    Explanations

    terms related to patient safety and healthcare risk management

    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
    307
    +0.38
    2.3%
    306
    +0.19
    1.1%
    269
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    307
    +0.38
    0.15
    418
    +0.19
    0.04
    306
    +0.11
    0.05
    Negative Logits
     defence
    -2.35
     offence
    -2.32
     flavour
    -2.05
     colour
    -2.00
     colours
    -1.91
     Defence
    -1.79
     humour
    -1.78
     mould
    -1.76
     organisation
    -1.75
    ’.
    -1.65
    POSITIVE LOGITS
    ľĵ
    3.87
    Ī
    3.84
    ī
    3.84
    į
    3.83
    ĥ½
    3.83
    ↵        ↵
    3.82
    ↵↵
    3.82
    ↵
    3.82
    ↵↵
    3.82
    3.82
    Activations Density 2.890%

    No Known Activations