© 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 EvalsExportsinterp-engineNEWAPI 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. 23901
    Prev
    Next
    INDEX
    Explanations

    quoted strings or text within code-related contexts

    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
    87
    +0.10
    0.5%
    317
    +0.10
    0.5%
    27
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    177
    +0.10
    0.03
    318
    +0.10
    0.02
    188
    +0.10
    0.02
    Negative Logits
    sworth
    -1.53
     negligence
    -1.50
     INDIRECT
    -1.49
    cester
    -1.49
     DAMAGE
    -1.48
     peroxidase
    -1.48
     neglect
    -1.47
     refrain
    -1.45
     NEGLIGENCE
    -1.44
     forgive
    -1.41
    POSITIVE LOGITS
    ģ
    2.10
    ĻĤ
    2.07
    IJ
    2.03
    ħ
    1.91
    dy
    1.76
    ¦
    1.75
    º
    1.75
    ı
    1.74
    ĩ
    1.72
    İ
    1.70
    Activations Density 0.164%

    No Known Activations