© 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. MLP Post
    9. 0-MLP-SM
    11. 26105
    Prev
    Next
    INDEX
    Explanations

    the word "mal" indicating a reference to something negative or harmful

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.0.hook_mlp_out
    Hook Layer
    0
    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
    148
    +0.13
    0.7%
    376
    +0.11
    0.6%
    255
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    469
    +0.13
    0.01
    303
    +0.11
    0.01
    255
    +0.11
    0.01
    Negative Logits
    ·¸
    -2.12
    ĸ´
    -2.09
    »¿
    -2.06
    ´
    -1.88
    ¿½
    -1.86
    Ĥ¬
    -1.81
    ij
    -1.75
    Ĥ
    -1.74
    ķ
    -1.72
    ĥ½
    -1.72
    POSITIVE LOGITS
    ieu
    1.83
    ieux
    1.79
    films
    1.78
    lege
    1.77
    don
    1.76
    lee
    1.74
    colm
    1.71
    fest
    1.69
    leys
    1.66
    1024
    1.59
    Activations Density 0.014%

    No Known Activations