© 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. 25346
    Prev
    Next
    INDEX
    Explanations

    terms related to categorization and classification

    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
    156
    +0.13
    0.7%
    203
    +0.13
    0.7%
    464
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.13
    0.07
    78
    +0.13
    0.05
    299
    +0.12
    0.06
    Negative Logits
    ping
    -1.79
    ifying
    -1.79
    erving
    -1.67
    âĢ¦]
    -1.60
     ours
    -1.58
    tering
    -1.57
    ing
    -1.54
    bing
    -1.53
    hering
    -1.50
     itself
    -1.49
    POSITIVE LOGITS
    ĻĤ
    3.02
    ĨĴ
    2.65
    ¿½
    2.53
    ↵
    2.45
    2.45
    2.45
    ↵
    2.45
    ↵
    2.45
     ↵
    2.45
    ↵↵
    2.45
    Activations Density 2.122%

    No Known Activations