© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 26955
    Prev
    Next
    INDEX
    Explanations

    instances of copyright years in the document

    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
    330
    +0.12
    0.7%
    497
    +0.11
    0.6%
    223
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    330
    +0.12
    0.03
    223
    +0.11
    0.02
    424
    +0.11
    0.02
    Negative Logits
     "</
    -1.57
     naked
    -1.43
    rvert
    -1.42
     jeans
    -1.42
     shorts
    -1.38
    tract
    -1.36
     tutorials
    -1.36
     enroll
    -1.35
    formaldehyde
    -1.35
    ()));
    -1.34
    POSITIVE LOGITS
     Kong
    1.61
    ]:
    1.59
    arde
    1.56
    astic
    1.55
    obia
    1.52
    aus
    1.49
    bol
    1.47
    pora
    1.46
     Crime
    1.44
    oes
    1.41
    Activations Density 0.040%

    No Known Activations