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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    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. 28912
    Prev
    Next
    INDEX
    Explanations

    references to dietary supplements

    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
    29
    +0.12
    0.6%
    217
    +0.10
    0.6%
    262
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    86
    +0.12
    0.04
    253
    +0.10
    0.04
    38
    +0.10
    0.03
    Negative Logits
    ®
    -2.92
    ĨĴ
    -2.86
    Ĵ
    -2.68
    ı
    -2.52
    ¸
    -2.50
    ·
    -2.44
    Ļª
    -2.40
    ½
    -2.35
    »
    -2.30
    ¶
    -2.27
    POSITIVE LOGITS
    lemental
    1.95
    cknowled
    1.95
    ellation
    1.72
    alties
    1.71
    ESSION
    1.68
    ation
    1.67
    teral
    1.63
    holder
    1.62
    eness
    1.61
    holders
    1.59
    Activations Density 0.666%

    No Known Activations