© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Joseph Bloom · Initial Gemma-2B SAEs
    3. Gemma-2B
    4. Residual Stream
    5. 0-RES-JB
    6. 1083
    Prev
    Next
    INDEX
    Explanations

    text related to returning items and contacting customer service

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.0.hook_resid_post_16384_anthropic
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    HuggingFaceFW/fineweb
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    690
    +0.06
    0.3%
    1699
    +0.06
    0.3%
    1871
    +0.06
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    906
    +0.06
    0.10
    1120
    +0.06
    0.11
    783
    +0.06
    0.09
    Negative Logits
    <bos>
    -1.78
    <?
    -0.99
    ⓧ
    -0.88
    
    
    -0.85
    <?
    
    -0.84
    /**
    -0.83
    /*
    -0.77
     reinstate
    -0.75
     expand
    -0.75
     maintain
    -0.74
    POSITIVE LOGITS
     vété
    1.95
     Juf
    1.92
     effe
    1.86
     ftu
    1.84
     stockholm
    1.84
     wien
    1.84
     aen
    1.83
     fta
    1.82
     „,
    1.82
     meis
    1.82
    Activations Density 1.115%

    No Known Activations