© 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. 30392
    Prev
    Next
    INDEX
    Explanations

    HTML tags and attributes related to web links and resources

    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
    84
    +0.11
    0.6%
    320
    +0.11
    0.6%
    74
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    74
    +0.11
    0.02
    37
    +0.11
    0.01
    379
    +0.10
    0.02
    Negative Logits
    àµį
    -2.00
     robbery
    -1.65
     robber
    -1.58
     conversion
    -1.57
    àµ
    -1.55
    ()</
    -1.53
     killer
    -1.52
     robbed
    -1.49
    alities
    -1.48
     theft
    -1.48
    POSITIVE LOGITS
    http
    1.52
    chel
    1.47
    orum
    1.45
    ©
    1.43
    uds
    1.43
     oc
    1.42
    ±
    1.40
    ³
    1.38
    orld
    1.36
    ="../../../../../../
    1.36
    Activations Density 0.090%

    No Known Activations