© 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. 9771
    Prev
    Next
    INDEX
    Explanations

    references to rats in various 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
    156
    +0.22
    1.3%
    30
    +0.15
    0.9%
    321
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    30
    +0.22
    0.01
    389
    +0.15
    0.01
    73
    +0.11
    0.01
    Negative Logits
     {{
    -1.55
     fact
    -1.53
    ferentially
    -1.51
    >$
    -1.49
     Appellee
    -1.48
    >>>
    -1.47
     semiconductor
    -1.44
    ...]
    -1.42
    lection
    -1.36
    visibility
    -1.35
    POSITIVE LOGITS
    ģ
    2.94
    Ģ
    2.66
    º
    2.62
    ĻĤ
    2.60
    ħ
    2.59
    ²
    2.39
    ŀ
    2.29
    Į
    2.29
    ¢
    2.28
    ¾
    2.27
    Activations Density 0.042%

    No Known Activations