Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 28903
    Prev
    Next
    INDEX
    Explanations

    terms related to various forms of artistic expression and their creators

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-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_resid_post
    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
    156
    +0.22
    1.3%
    410
    +0.14
    0.8%
    407
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    365
    +0.22
    0.05
    222
    +0.14
    0.04
    321
    +0.13
    0.04
    Negative Logits
    ĥ½
    -1.65
     DISTRICT
    -1.62
    Ħ
    -1.56
    urbs
    -1.49
    iffer
    -1.44
     Legislature
    -1.42
    orb
    -1.41
    IBLE
    -1.37
     Debtors
    -1.36
     Patent
    -1.36
    POSITIVE LOGITS
    ship
    2.42
    ships
    2.23
    liness
    2.10
     living
    1.87
    himself
    1.82
    hood
    1.75
    friends
    1.73
     himself
    1.67
    acing
    1.65
    ily
    1.62
    Activations Density 0.435%

    No Known Activations