© 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. 1-RES-SM
    11. 29651
    Prev
    Next
    INDEX
    Explanations

    instances of the word "down" and its variations

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    248
    +0.16
    0.9%
    414
    +0.12
    0.6%
    296
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    248
    +0.16
    0.06
    414
    +0.12
    0.05
    198
    +0.11
    0.03
    Negative Logits
    reland
    -1.95
    Ī
    -1.83
    ĥ½
    -1.69
    US
    -1.56
    ½
    -1.56
     independence
    -1.54
    ķ
    -1.48
     always
    -1.48
    dern
    -1.44
    rena
    -1.43
    POSITIVE LOGITS
    grades
    2.00
    graded
    1.84
    bound
    1.84
    river
    1.80
    loaded
    1.68
    tics
    1.61
    ceptors
    1.60
    pez
    1.58
    ductory
    1.57
    gart
    1.56
    Activations Density 0.070%

    No Known Activations