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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 29465
    Prev
    Next
    INDEX
    Explanations

    percentage values and their context in data discussions

    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
    186
    +0.15
    0.9%
    288
    +0.15
    0.8%
    161
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    187
    +0.15
    0.07
    288
    +0.15
    0.13
    7
    +0.11
    -0.01
    Negative Logits
    Ĵ
    -2.01
    Ĥ
    -1.98
    ¸
    -1.70
    ·¸
    -1.69
    Ľ
    -1.65
    »
    -1.62
    iendo
    -1.61
    Ĭ
    -1.59
    Ģ
    -1.58
    ág
    -1.56
    POSITIVE LOGITS
    à¯ģ
    1.59
    pired
    1.50
    ?’
    1.45
    ி
    1.42
    ?'
    1.41
    gran
    1.37
    á¼
    1.37
     loyalty
    1.31
    ????
    1.30
     ?
    1.27
    Activations Density 4.453%

    No Known Activations