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. 27336
    Prev
    Next
    INDEX
    Explanations

    references to community and group identity

    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
    195
    +0.12
    0.7%
    394
    +0.12
    0.7%
    465
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    338
    +0.12
    0.12
    402
    +0.12
    0.12
    114
    +0.11
    0.09
    Negative Logits
    ĥ½
    -4.43
    »¿
    -4.34
    »
    -3.64
    ¡
    -3.62
    ¬
    -3.60
    Īĺ
    -3.58
    ·¸
    -3.51
    ĩ
    -3.43
    ĥ
    -3.41
    ³
    -3.39
    POSITIVE LOGITS
     coffin
    1.65
    eness
    1.57
     further
    1.51
     otherwise
    1.47
    matter
    1.45
    .;
    1.43
     wedge
    1.43
     bones
    1.42
     stub
    1.34
     skeleton
    1.33
    Activations Density 0.279%

    No Known Activations