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. 27590
    Prev
    Next
    INDEX
    Explanations

    the term "gall," potentially indicating a focus on discussions regarding resilience or boldness

    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
    300
    +0.14
    0.8%
    429
    +0.14
    0.8%
    349
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1
    +0.14
    0.01
    260
    +0.14
    0.01
    349
    +0.13
    0.01
    Negative Logits
    )){
    -1.84
     slightest
    -1.62
     cens
    -1.61
     otherwise
    -1.50
     certain
    -1.49
    ores
    -1.46
    cens
    -1.43
    arer
    -1.42
    ges
    -1.42
     priori
    -1.41
    POSITIVE LOGITS
    stones
    2.12
    ©
    2.07
    ery
    1.97
    ¢
    1.77
    phrase
    1.72
    ERY
    1.70
    stone
    1.64
    ableView
    1.62
    aho
    1.61
    ¾
    1.60
    Activations Density 0.020%

    No Known Activations