© 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. 20772
    Prev
    Next
    INDEX
    Explanations

    verbs that involve halting or preventing something

    oai_token-act-pair · gpt-3.5-turbo

    repeated calls to take action or halt negative behaviors

    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
    Sov
    -0.82
    ammy
    -0.77
    ighth
    -0.75
    olesc
    -0.74
    ault
    -0.71
    aths
    -0.69
    ety
    -0.69
    VERTISEMENT
    -0.69
    uth
    -0.67
    orthy
    -0.67
    POSITIVE LOGITS
    gap
    0.95
    watching
    0.93
    watch
    0.93
     bothering
    0.86
     stopping
    0.76
     breathing
    0.75
     smoking
    0.75
    reon
    0.75
     blinking
    0.73
    door
    0.71
    Activations Density 0.025%

    No Known Activations