© 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
    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. 9461
    Prev
    Next
    INDEX
    Explanations

    instances of phrases related to publicly expressing an opinion or stance

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "came out" or variations of it

    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
    olesterol
    -0.67
    nesota
    -0.66
    »Ĵ
    -0.63
    ille
    -0.62
     Bots
    -0.61
    Login
    -0.61
    cius
    -0.60
     Icar
    -0.59
    =-=-=-=-=-=-=-=-
    -0.58
    atomic
    -0.58
    POSITIVE LOGITS
    fitted
    1.05
    doors
    0.83
     swinging
    0.83
    wards
    0.82
    smart
    0.79
    door
    0.75
    board
    0.75
    rower
    0.73
     stronger
    0.72
    skirts
    0.72
    Activations Density 0.040%

    No Known Activations