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 SAEs
    3. Gemma-2B
    4. Residual Stream
    5. 0-RES-JB
    6. 39
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "drone"

    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
    1757
    +0.12
    0.4%
    1323
    +0.09
    0.3%
    1052
    +0.08
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    690
    +0.12
    0.04
    1343
    +0.09
    0.04
    1363
    +0.08
    0.03
    Negative Logits
    public
    -0.76
    class
    -0.76
    itemize
    -0.74
                
    -0.73
    -0.73
    <bos>
    -0.73
        
    -0.72
    private
    -0.72
    	
    -0.71
    for
    -0.71
    POSITIVE LOGITS
     maneu
    2.38
     milf
    2.25
     increa
    2.21
     impra
    2.20
     strick
    2.20
     hentai
    2.17
     affor
    2.17
     disreg
    2.15
     reluct
    2.15
     emphat
    2.11
    Activations Density 0.338%

    No Known Activations