© 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. 2-RES-SM
    11. 10296
    Prev
    Next
    INDEX
    Explanations

    references to self-awareness and introspection

    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
    369
    +0.14
    0.8%
    164
    +0.14
    0.8%
    484
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    279
    +0.14
    0.03
    77
    +0.14
    0.01
    410
    +0.12
    0.03
    Negative Logits
     ICU
    -1.53
     course
    -1.49
    wagen
    -1.39
    align
    -1.38
     NW
    -1.38
    ition
    -1.38
     notation
    -1.36
    uary
    -1.33
    CDATA
    -1.32
    collapse
    -1.32
    POSITIVE LOGITS
    esome
    2.33
    orld
    2.20
    areness
    2.00
    ful
    1.75
    lessly
    1.73
    indo
    1.73
    nee
    1.70
    eston
    1.69
    yer
    1.63
    yers
    1.61
    Activations Density 0.190%

    No Known Activations