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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 26425
    Prev
    Next
    INDEX
    Explanations

    instances of legal terminology and concepts related to pre-emption and jurisdiction

    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
    206
    +0.15
    0.9%
    301
    +0.13
    0.7%
    200
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    200
    +0.15
    0.07
    23
    +0.13
    0.04
    74
    +0.13
    0.03
    Negative Logits
    legraph
    -1.70
    posium
    -1.58
    pmed
    -1.58
    pgen
    -1.56
    ↵ ↵
    -1.54
    ↵
    -1.50
     via
    -1.50
    ↵ ↵
    -1.48
    pany
    -1.43
    endar
    -1.39
    POSITIVE LOGITS
    '?"
    1.72
    points
    1.55
     manners
    1.52
     rats
    1.46
     assumptions
    1.45
    ?'
    1.45
     coworkers
    1.45
     Customers
    1.41
    ?_
    1.41
    '?
    1.39
    Activations Density 2.353%

    No Known Activations