© 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. Residuals
    9. 2-RES-SM
    11. 28632
    Prev
    Next
    INDEX
    Explanations

    terms related to software licenses and permissions

    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
    71
    +0.12
    0.7%
    369
    +0.12
    0.7%
    74
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    74
    +0.12
    0.02
    222
    +0.12
    0.02
    86
    +0.11
    0.02
    Negative Logits
    ľĵ
    -2.35
    Ĺ
    -1.86
     purpose
    -1.62
    ÑĢоÑģ
    -1.57
    purpose
    -1.57
    stood
    -1.55
    Į
    -1.50
    opsies
    -1.50
    Ł
    -1.50
    >",
    -1.48
    POSITIVE LOGITS
    uelle
    1.61
    umab
    1.50
    itious
    1.48
     religion
    1.47
    aceous
    1.47
     den
    1.46
    bery
    1.45
    holm
    1.42
     yourselves
    1.39
    ner
    1.38
    Activations Density 0.197%

    No Known Activations