© 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. 24876
    Prev
    Next
    INDEX
    Explanations

    instances of the word "contents."

    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
    500
    +0.16
    0.9%
    392
    +0.13
    0.7%
    409
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    500
    +0.16
    0.03
    120
    +0.13
    0.06
    392
    +0.10
    0.04
    Negative Logits
    ĻĤ
    -3.35
    ĵ
    -3.25
    ¤
    -3.15
    ¨
    -3.11
    ģ
    -3.06
    Ĵ
    -3.00
    ŀ
    -2.98
    IJ
    -2.96
    ĺ
    -2.96
    Ļ
    -2.92
    POSITIVE LOGITS
    INGTON
    1.58
    wards
    1.50
    ffect
    1.37
     stain
    1.36
    atives
    1.35
     brightest
    1.35
     detect
    1.33
    arding
    1.32
    atin
    1.30
    ="'
    1.30
    Activations Density 0.767%

    No Known Activations