Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 29033
    Prev
    Next
    INDEX
    Explanations

    conditional statements expressing hypothetical scenarios

    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
    258
    +0.14
    0.8%
    478
    +0.13
    0.7%
    486
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    486
    +0.14
    0.05
    236
    +0.13
    0.06
    481
    +0.13
    0.05
    Negative Logits
    »¿
    -2.60
    Ļª
    -2.42
    ĻĤ
    -2.27
    ĩ
    -2.26
    Ĩ
    -2.08
    ĭ
    -2.08
    ľĵ
    -2.04
    ¡
    -2.03
    ¼
    -2.00
    Ħ
    -1.97
    POSITIVE LOGITS
    iox
    1.95
    icar
    1.61
     duplicate
    1.53
     untimely
    1.48
    zia
    1.43
     dye
    1.42
     sacrifice
    1.41
     prob
    1.39
    ffen
    1.38
    ismiss
    1.36
    Activations Density 0.189%

    No Known Activations