© 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 EvalsExportsinterp-engineNEWAPI 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. 24240
    Prev
    Next
    INDEX
    Explanations

    the word "clearly" and its variations to indicate strong assertions or observations

    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
    23
    +0.18
    1.0%
    376
    +0.15
    0.8%
    377
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.18
    0.07
    156
    +0.15
    0.04
    298
    +0.12
    0.05
    Negative Logits
    aurus
    -2.01
    Ķ
    -1.88
    Ħ
    -1.86
    ħ
    -1.77
    Ľ
    -1.75
    Ģ
    -1.72
    pere
    -1.68
    Ŀ
    -1.63
    §
    -1.60
    AMS
    -1.54
    POSITIVE LOGITS
     identifiable
    1.54
    acted
    1.49
     rect
    1.48
    ffected
    1.44
    fix
    1.44
     erroneous
    1.43
     ascertained
    1.43
     enough
    1.43
    limits
    1.41
     POSS
    1.40
    Activations Density 0.378%

    No Known Activations