© 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. 6946
    Prev
    Next
    INDEX
    Explanations

    Interestingly, Neuron 4 appears to be looking for the word "Dis" which seems to be a prefix indicating some kind of negative connotation or disruption

    oai_token-act-pair · gpt-3.5-turbo

    words or phrases that denote various forms or disciplines of behavior or compliance

    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
    hetti
    -0.70
    glers
    -0.68
    OPLE
    -0.67
     Kinnikuman
    -0.64
    imately
    -0.63
     Scotia
    -0.62
     Juliet
    -0.62
     Metatron
    -0.62
     Osw
    -0.61
     Hanna
    -0.61
    POSITIVE LOGITS
    cipl
    1.21
    claimer
    1.19
    ruption
    1.19
    qus
    1.14
    comfort
    1.14
    rup
    1.14
    cerning
    1.12
    enfranch
    1.11
    ciples
    1.11
    burse
    1.10
    Activations Density 0.020%

    No Known Activations