© 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. 653
    Prev
    Next
    INDEX
    Explanations

    negative statements followed by a strong contradiction or rebuttal

    oai_token-act-pair · gpt-3.5-turbo

    instances of negation or denial in statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    expressions of denial or negation starting with "No, this is/that is not."

    oai_token-act-pair · gpt-4oTriggered by @tatsatx
    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
     PID
    -0.75
    abal
    -0.72
     tro
    -0.71
    gypt
    -0.65
    espie
    -0.61
    ounded
    -0.59
     redes
    -0.58
     culminated
    -0.58
     paced
    -0.57
     trave
    -0.57
    POSITIVE LOGITS
     sir
    0.92
    nor
    0.83
    onsense
    0.79
     nor
    0.77
    æĪ
    0.74
    tons
    0.73
     dear
    0.71
    Nor
    0.70
     Nor
    0.68
    stop
    0.67
    Activations Density 0.070%

    No Known Activations