© 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
    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. 17593
    Prev
    Next
    INDEX
    Explanations

    modifiers indicating a high degree of an attribute or action

    oai_token-act-pair · gpt-3.5-turbo

    intensifiers used to emphasize qualities or characteristics

    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
    enhagen
    -0.75
    iem
    -0.71
     Travels
    -0.69
     Pavilion
    -0.68
    RY
    -0.66
    DAQ
    -0.65
    Frames
    -0.65
    antry
    -0.65
    éŃĶ
    -0.64
    esi
    -0.63
    POSITIVE LOGITS
    bad
    0.97
     bad
    0.95
     BAD
    0.92
     nasty
    0.88
     cool
    0.88
     GOOD
    0.88
     naughty
    0.87
     juicy
    0.86
     shitty
    0.85
    nice
    0.84
    Activations Density 0.108%

    No Known Activations