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

    Neuronpedia

    Jacobian LensNEW
    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. 25077
    Prev
    Next
    INDEX
    Explanations

    references to groups or collective entities

    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
    157
    +0.13
    0.7%
    446
    +0.12
    0.7%
    424
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    361
    +0.13
    0.04
    339
    +0.12
    0.03
    488
    +0.10
    0.03
    Negative Logits
    respectively
    -1.53
     direction
    -1.52
     respectively
    -1.51
    iation
    -1.51
     currents
    -1.51
     together
    -1.45
     similarly
    -1.43
     relevant
    -1.40
     independ
    -1.39
    eurs
    -1.36
    POSITIVE LOGITS
    ī
    2.17
    ·
    1.99
    ¯
    1.98
    Ĥ¬
    1.91
    Į
    1.90
    ·¸
    1.89
    ¥
    1.89
    ↵
    1.81
    1.81
    ↵↵
    1.81
    Activations Density 0.446%

    No Known Activations