© 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
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 10377
    Prev
    Next
    INDEX
    Explanations

    phrases related to position or location, specifically referring to being situated behind something

    oai_token-act-pair · gpt-3.5-turbo

    the word "behind" and its variations in different contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    istic
    -0.81
     Pwr
    -0.75
    iser
    -0.70
    olid
    -0.67
    ISM
    -0.66
    ogl
    -0.65
    acci
    -0.65
    imer
    -0.65
    izens
    -0.64
    ERC
    -0.63
    POSITIVE LOGITS
    standing
    0.74
    shore
    0.74
    coat
    0.71
    tions
    0.69
    lined
    0.69
    gression
    0.68
     bars
    0.68
     stump
    0.67
    ¬
    0.66
    side
    0.66
    Activations Density 0.017%

    No Known Activations