© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 1-RES-SM
    11. 27960
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate intent or purpose related to actions

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    94
    +0.12
    0.7%
    23
    +0.11
    0.6%
    402
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.12
    0.07
    478
    +0.11
    0.00
    505
    +0.11
    0.03
    Negative Logits
     channels
    -1.67
     sheets
    -1.59
    uscript
    -1.59
     lanes
    -1.54
    ressions
    -1.52
     pages
    -1.50
     Caption
    -1.48
    resses
    -1.46
     Statutes
    -1.45
    nai
    -1.44
    POSITIVE LOGITS
    Ĺ
    2.65
    ¦
    2.64
    ĨĴ
    2.53
    ·
    2.52
    »
    2.49
    ©
    2.40
    ·¸
    2.38
    Į
    2.33
    ²
    2.29
    ¶
    2.28
    Activations Density 1.356%

    No Known Activations