© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 15110
    Prev
    Next
    INDEX
    Explanations

    negative adjectives

    oai_token-act-pair · gpt-3.5-turbo

    negative adjectives or phrases that convey criticism or disdain

    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
    HOME
    -0.59
     waivers
    -0.51
     Rhino
    -0.50
     FB
    -0.50
     restrooms
    -0.49
     preferably
    -0.49
     webcam
    -0.47
     cycle
    -0.47
    hyde
    -0.47
     feds
    -0.47
    POSITIVE LOGITS
    ious
    0.97
    iously
    0.94
    acious
    0.92
    itud
    0.92
    entious
    0.92
    arious
    0.92
    isive
    0.90
    assion
    0.88
    uous
    0.87
    itatively
    0.87
    Activations Density 0.427%

    No Known Activations