© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B SAEs
    5. Gemma-2B
    7. Residual Stream
    9. 0-RES-JB
    11. 1511
    Prev
    Next
    INDEX
    Explanations

    terms related to justifications, ethics, and decision-making

    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
    1896
    +0.07
    0.2%
    899
    +0.07
    0.2%
    370
    +0.07
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    859
    +0.07
    0.02
    1793
    +0.07
    0.02
    899
    +0.07
    0.02
    Negative Logits
    <bos>
    -0.76
    発
    -0.60
     forward
    -0.57
    NewReader
    -0.57
    public
    -0.56
     noon
    -0.56
    arr
    -0.56
     either
    -0.55
     launched
    -0.55
     لكرة
    -0.54
    POSITIVE LOGITS
     justified
    2.42
     justify
    2.33
     justifying
    2.27
     justifies
    2.26
     justification
    2.21
     gius
    2.12
     justific
    1.98
     Justification
    1.94
    justified
    1.91
     jus
    1.89
    Activations Density 0.133%

    No Known Activations