© 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. 29633
    Prev
    Next
    INDEX
    Explanations

    the presence of editorial content and references to editors

    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
    +0.13
    0.8%
    410
    +0.13
    0.8%
    187
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.13
    0.03
    253
    +0.13
    0.02
    77
    +0.12
    0.01
    Negative Logits
     proper
    -1.79
     populated
    -1.54
     else
    -1.49
     ago
    -1.48
     hereafter
    -1.45
    ClickListener
    -1.44
     true
    -1.42
     longer
    -1.42
    stic
    -1.39
    ]{}]{}
    -1.36
    POSITIVE LOGITS
    º
    3.13
    «
    2.84
    ¾
    2.83
    Ń
    2.82
    §
    2.77
    ļ
    2.71
    ĻĤ
    2.67
    Ĵ
    2.66
    ¦
    2.64
    ł
    2.63
    Activations Density 0.143%

    No Known Activations