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. 29233
    Prev
    Next
    INDEX
    Explanations

    legal terms and components related to court cases

    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
    410
    +0.15
    0.9%
    21
    +0.12
    0.7%
    104
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    21
    +0.15
    0.03
    168
    +0.12
    0.03
    367
    +0.12
    0.02
    Negative Logits
     depends
    -1.72
    cribe
    -1.67
    itary
    -1.65
     varies
    -1.60
    eless
    -1.52
    ocracy
    -1.47
    tons
    -1.46
    imate
    -1.46
    avirus
    -1.40
    Notify
    -1.40
    POSITIVE LOGITS
    Īĺ
    3.13
                                            
    2.86
    ↵Č      
    2.86
    ↵      
    2.86
     ↵   
    2.86
    č↵       
    2.86
                                                                              
    2.86
     ↵     
    2.86
    <|outofrange|>
    2.86
    č↵    
    2.86
    Activations Density 0.093%

    No Known Activations