© 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. 2-RES-SM
    11. 12347
    Prev
    Next
    INDEX
    Explanations

    the word "super" in various contexts

    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
    +0.15
    0.9%
    280
    +0.12
    0.7%
    198
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    511
    +0.15
    0.02
    386
    +0.12
    0.01
    446
    +0.12
    0.02
    Negative Logits
    uximab
    -1.94
    rapeut
    -1.67
    EMENT
    -1.66
    estinal
    -1.58
    abouts
    -1.56
    RICT
    -1.54
    izumab
    -1.52
    ês
    -1.52
    ually
    -1.50
    OOGLE
    -1.49
    POSITIVE LOGITS
    ¼
    1.94
    visor
    1.91
    visors
    1.85
    Ĥ
    1.84
    conduct
    1.82
    visory
    1.70
    Ł
    1.62
    conductor
    1.62
    µ
    1.56
    formance
    1.54
    Activations Density 0.114%

    No Known Activations