© 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. 31640
    Prev
    Next
    INDEX
    Explanations

    phrases expressing the degree of possibility or potential in a scenario

    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
    176
    +0.15
    0.9%
    499
    +0.13
    0.8%
    478
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    176
    +0.15
    0.05
    91
    +0.13
    0.04
    193
    +0.11
    0.04
    Negative Logits
    rium
    -1.70
    holder
    -1.58
    staff
    -1.51
    hold
    -1.51
    oster
    -1.45
    ij
    -1.45
     officer
    -1.44
    atted
    -1.40
    etary
    -1.39
    lich
    -1.36
    POSITIVE LOGITS
     possible
    1.94
    possible
    1.77
    gard
    1.73
    inks
    1.64
    phalt
    1.58
    pect
    1.56
    thy
    1.48
    ides
    1.44
    gaard
    1.42
    cedes
    1.40
    Activations Density 0.199%

    No Known Activations