© 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. 0-MLP-SM
    11. 30786
    Prev
    Next
    INDEX
    Explanations

    the word "when" in various 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/0-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.0.hook_mlp_out
    Hook Layer
    0
    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
    446
    +0.15
    0.8%
    462
    +0.13
    0.7%
    307
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    174
    +0.15
    0.07
    154
    +0.13
    0.08
    112
    +0.12
    0.07
    Negative Logits
    fiddle
    -2.09
    ħ
    -1.98
    Īĺ
    -1.88
    ¶
    -1.75
    ĥ
    -1.73
    cha
    -1.72
    IJ
    -1.70
    Ł
    -1.69
    ³
    -1.67
    ¿½
    -1.66
    POSITIVE LOGITS
    rap
    1.63
    respectively
    1.50
    quez
    1.42
     lapar
    1.40
     norm
    1.39
    orems
    1.39
    fact
    1.36
    resa
    1.34
     respectively
    1.30
    prototype
    1.30
    Activations Density 0.098%

    No Known Activations