© 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. 28973
    Prev
    Next
    INDEX
    Explanations

    various methods or strategies related to problem-solving

    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
    376
    +0.12
    0.7%
    172
    +0.11
    0.6%
    117
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    117
    +0.12
    0.02
    172
    +0.11
    0.02
    433
    +0.10
    0.02
    Negative Logits
    »¿
    -2.65
     oath
    -2.06
    ĥ½
    -1.99
    ½
    -1.97
    ¿½
    -1.72
    ¾
    -1.69
    ®
    -1.68
    Ĭ
    -1.62
    uvre
    -1.59
     hereby
    -1.57
    POSITIVE LOGITS
    able
    1.94
    gun
    1.83
    iles
    1.61
     targeting
    1.59
    points
    1.56
    finding
    1.53
     toward
    1.51
    ologies
    1.50
     objectives
    1.50
    nikov
    1.49
    Activations Density 0.086%

    No Known Activations