© 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. 28236
    Prev
    Next
    INDEX
    Explanations

    words or phrases suggesting negative or critical 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/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
    207
    +0.13
    0.7%
    497
    +0.11
    0.6%
    507
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    249
    +0.13
    0.04
    203
    +0.11
    0.03
    497
    +0.10
    0.02
    Negative Logits
    ```
    -1.83
    ="$(
    -1.69
    ynchron
    -1.62
     %>
    -1.62
    romycin
    -1.55
    CTYPE
    -1.50
    ]):
    -1.48
    :&
    -1.48
    yg
    -1.45
    chemia
    -1.42
    POSITIVE LOGITS
     himself
    1.81
    backs
    1.77
     herself
    1.65
     commissioned
    1.53
    eur
    1.47
    boards
    1.43
    conscious
    1.43
    pent
    1.39
     storyline
    1.38
    room
    1.37
    Activations Density 0.191%

    No Known Activations