© 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. 27719
    Prev
    Next
    INDEX
    Explanations

    categories and organization labels in texts

    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
    93
    +0.14
    0.8%
    170
    +0.13
    0.7%
    507
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    93
    +0.14
    0.04
    300
    +0.13
    0.03
    374
    +0.12
    0.01
    Negative Logits
    iors
    -1.95
    ĻĤ
    -1.92
    ł
    -1.72
    «
    -1.66
    Īĺ
    -1.65
     weights
    -1.54
     Officer
    -1.46
    ĸ
    -1.45
     note
    -1.41
    ģ
    -1.41
    POSITIVE LOGITS
    etically
    1.63
    olith
    1.56
    efore
    1.54
    à±ģ
    1.51
     epit
    1.47
    stones
    1.44
    ymph
    1.43
    etica
    1.40
    ascus
    1.39
    inous
    1.38
    Activations Density 0.146%

    No Known Activations