© 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. 2-MLP-SM
    11. 23301
    Prev
    Next
    INDEX
    Explanations

    repetitive phrases, particularly the word "for" used in varying contexts

    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/2-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.2.hook_mlp_out
    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
    123
    +0.17
    1.0%
    54
    +0.15
    0.8%
    500
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    123
    +0.17
    0.10
    178
    +0.15
    0.09
    500
    +0.11
    0.07
    Negative Logits
    ème
    -1.57
    ubation
    -1.56
    arn
    -1.55
     couple
    -1.54
     bunch
    -1.46
     Leib
    -1.44
    ÃŁe
    -1.43
    ights
    -1.42
    e
    -1.41
    ef
    -1.35
    POSITIVE LOGITS
    ŀ
    1.82
    Ļª
    1.77
    ĥ½
    1.70
    journals
    1.70
    idopsis
    1.60
    ĨĴ
    1.59
    blogger
    1.57
    ħ
    1.55
    ↵
    1.52
    č↵
    1.52
    Activations Density 0.635%

    No Known Activations