© 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. 25568
    Prev
    Next
    INDEX
    Explanations

    the word "Although."

    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
    23
    +0.19
    1.1%
    376
    +0.16
    0.9%
    210
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    135
    +0.19
    0.03
    508
    +0.16
    0.03
    210
    +0.13
    0.03
    Negative Logits
    ipore
    -1.92
    ented
    -1.70
    ”.
    -1.70
    earchers
    -1.66
    enium
    -1.60
    itud
    -1.53
    abad
    -1.53
    atories
    -1.53
    ”:
    -1.52
    ated
    -1.51
    POSITIVE LOGITS
    ¬
    1.89
     bond
    1.69
    IJ
    1.59
    Ģ
    1.58
    ¸
    1.58
    ¤
    1.56
    ¶
    1.48
    noon
    1.47
    ¹
    1.45
     borrowing
    1.35
    Activations Density 3.587%

    No Known Activations