Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. MLP Post
    5. 0-MLP-SM
    6. 27298
    Prev
    Next
    INDEX
    Explanations

    code elements related to functions and return statements

    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
    115
    +0.11
    0.6%
    510
    +0.11
    0.6%
    366
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    366
    +0.11
    0.07
    342
    +0.11
    0.07
    197
    +0.11
    0.06
    Negative Logits
    ·
    -1.96
    Ļ
    -1.73
    ª
    -1.60
    °
    -1.60
    Ģ
    -1.60
    ´
    -1.60
    «
    -1.59
    ij
    -1.59
    µ
    -1.52
    activated
    -1.51
    POSITIVE LOGITS
    eLife
    1.77
    ificantly
    1.76
    reads
    1.59
     portrait
    1.48
    oracle
    1.43
     concer
    1.43
    hood
    1.40
     regards
    1.40
    nost
    1.38
    itives
    1.38
    Activations Density 0.558%

    No Known Activations