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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 26008
    Prev
    Next
    INDEX
    Explanations

    terms related to educational and developmental programs

    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
    135
    +0.12
    0.6%
    460
    +0.11
    0.6%
    63
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.12
    0.04
    491
    +0.11
    0.05
    276
    +0.11
    0.05
    Negative Logits
    fecture
    -1.79
    club
    -1.64
    union
    -1.63
    Ĩ
    -1.59
    heng
    -1.58
     won
    -1.46
     union
    -1.45
    rosse
    -1.45
    encoder
    -1.39
    decoder
    -1.37
    POSITIVE LOGITS
     boots
    1.66
     thickness
    1.60
     APPEALS
    1.51
    WHM
    1.45
    ologists
    1.44
     sunlight
    1.44
     gallons
    1.43
    OUR
    1.43
     someone
    1.39
     alternatives
    1.37
    Activations Density 1.783%

    No Known Activations