© 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. 4-RES-SM
    11. 13184
    Prev
    Next
    INDEX
    Explanations

    code-related statements and commands

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    397
    +0.32
    2.0%
    263
    +0.14
    0.9%
    78
    +0.12
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    9
    +0.32
    0.10
    397
    +0.14
    0.10
    451
    +0.12
    0.03
    Negative Logits
    aries
    -1.63
    oretic
    -1.44
    ator
    -1.40
     familiar
    -1.32
     annually
    -1.30
    irus
    -1.26
    ality
    -1.26
     closely
    -1.25
     everywhere
    -1.23
    onica
    -1.22
    POSITIVE LOGITS
    ¤
    2.58
    ĥ½
    2.57
    Ĥ
    2.48
    §
    2.41
    Ĥ¬
    2.40
    ²
    2.34
    ī
    2.31
    ľĵ
    2.31
    ¦
    2.29
    }
    2.23
    Activations Density 0.617%

    No Known Activations