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 · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 9693
    Prev
    Next
    INDEX
    Explanations

    the word "knock" and related variations

    oai_token-act-pair · gpt-3.5-turbo

    references to the word "knock" and its variations, indicating a focus on the concept of knocking or similar actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ional
    -0.79
    RED
    -0.72
    heid
    -0.72
    emale
    -0.70
    ivities
    -0.70
    rano
    -0.66
    ohyd
    -0.66
    mberg
    -0.66
    ranging
    -0.66
    uture
    -0.65
    POSITIVE LOGITS
    down
    1.12
     knock
    0.93
    downs
    0.83
     down
    0.83
    DOWN
    0.81
     unconscious
    0.80
    kn
    0.80
    ı
    0.78
    outs
    0.77
    ety
    0.75
    Activations Density 0.031%

    No Known Activations