© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 29884
    Prev
    Next
    INDEX
    Explanations

    references to medications or pharmaceutical terms

    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
    213
    +0.13
    0.7%
    211
    +0.12
    0.7%
    65
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    109
    +0.13
    0.00
    501
    +0.12
    0.00
    417
    +0.10
    0.00
    Negative Logits
     myself
    -1.97
     axes
    -1.64
     axis
    -1.64
     guys
    -1.60
    ist
    -1.59
    itta
    -1.59
     thumbs
    -1.55
     crossed
    -1.54
    ists
    -1.52
    elf
    -1.52
    POSITIVE LOGITS
    »¿
    1.60
    Applied
    1.60
    avor
    1.57
    ftware
    1.53
    à¯ģ
    1.47
    ounder
    1.46
    inib
    1.46
    ALSE
    1.42
    OMEM
    1.41
    aq
    1.40
    Activations Density 0.000%

    No Known Activations