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

    Neuronpedia

    Jacobian LensNEW
    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. 8-RES-JB
    11. 3891
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the action of leaving or being removed from a location or situation

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating the desire to leave or escape a situation

    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
    ieth
    -0.74
     bene
    -0.73
     futures
    -0.65
    audi
    -0.64
    ellen
    -0.63
    alian
    -0.62
     interpretations
    -0.61
     foresee
    -0.60
    icity
    -0.59
     succeeds
    -0.59
    POSITIVE LOGITS
     grips
    0.76
     habit
    0.73
     Truck
    0.70
    road
    0.65
     Dirt
    0.62
    eno
    0.62
     groove
    0.62
     speeding
    0.61
     Grip
    0.61
     grip
    0.61
    Activations Density 0.109%

    No Known Activations