© 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. 2-RES-SM
    11. 30355
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "defamation" and its variations in different contexts

    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
    133
    +0.12
    0.7%
    354
    +0.11
    0.6%
    177
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    415
    +0.12
    0.02
    386
    +0.11
    0.01
    367
    +0.10
    0.02
    Negative Logits
    ĨĴ
    -2.65
    ª
    -2.52
    ĩ
    -2.44
    Ī
    -2.21
    ¢
    -2.00
    Ķ
    -1.98
    »¿
    -1.96
    ı
    -1.96
    º
    -1.87
    ¬
    -1.84
    POSITIVE LOGITS
    initions
    2.60
    initely
    2.26
    ensively
    2.07
    unct
    1.67
    iciency
    1.67
    amation
    1.64
    ibr
    1.62
    irmat
    1.57
     Rather
    1.53
    icit
    1.52
    Activations Density 0.128%

    No Known Activations