© 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. 2-RES-SM
    11. 31221
    Prev
    Next
    INDEX
    Explanations

    mentions of patterns 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
    +0.14
    0.8%
    392
    +0.12
    0.7%
    169
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    392
    +0.14
    0.02
    169
    +0.12
    0.02
    429
    +0.12
    0.01
    Negative Logits
    ½
    -3.10
    ľĵ
    -2.96
    §
    -2.85
    «
    -2.81
    ¸
    -2.79
    ĥ½
    -2.69
    ¼
    -2.56
    ĸ´
    -2.49
    »¿
    -2.49
    ĻĤ
    -2.46
    POSITIVE LOGITS
    maker
    1.84
    etes
    1.73
    wise
    1.71
     emerges
    1.69
    ique
    1.66
    chart
    1.66
    ismatic
    1.64
    behind
    1.62
     occurs
    1.61
    atically
    1.60
    Activations Density 0.116%

    No Known Activations