© 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. 4-RES-SM
    11. 10874
    Prev
    Next
    INDEX
    Explanations

    negative connotations or critical evaluations related to various subjects

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    78
    +0.14
    0.8%
    412
    +0.11
    0.7%
    337
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    448
    +0.14
    0.03
    208
    +0.11
    0.03
    412
    +0.11
    0.03
    Negative Logits
    edly
    -1.79
    ophil
    -1.52
    iously
    -1.48
    ogens
    -1.39
    ily
    -1.39
     peroxidase
    -1.34
    arily
    -1.33
    atically
    -1.33
    ILY
    -1.32
     invariance
    -1.32
    POSITIVE LOGITS
    ³
    2.83
    ¹
    2.81
    ¼
    2.75
    »
    2.74
    ®
    2.69
    ¾
    2.67
    °
    2.66
    £
    2.63
    ´
    2.62
    ¡
    2.61
    Activations Density 0.093%

    No Known Activations