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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Initial Gemma-2B-IT SAEs
    3. Gemma-2B-IT
    4. Residual Stream
    5. 12-RES-JB
    6. 420
    Prev
    Next
    INDEX
    Explanations

    chemical compound synthesis processes

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.hook_resid_post
    Hook Layer
    12
    Architecture
    standard
    Context Size
    1,024
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    876
    +0.27
    0.9%
    1013
    +0.17
    0.6%
    283
    +0.15
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    876
    +0.27
    -0.00
    283
    +0.17
    0.01
    1499
    +0.15
    0.04
    Negative Logits
    <bos>
    -0.79
     Whigs
    -0.63
     guerrillas
    -0.58
     statesmen
    -0.58
     Saxons
    -0.57
    --)
    
    -0.57
     Puritans
    -0.56
    Père
    -0.55
     Confederacy
    -0.55
     boyhood
    -0.54
    POSITIVE LOGITS
     aen
    1.52
     waer
    1.51
     effe
    1.50
     ftu
    1.46
     meis
    1.42
     perfon
    1.42
     fup
    1.41
     vns
    1.37
     tranf
    1.36
     nece
    1.36
    Activations Density 0.147%

    No Known Activations