© 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. 10-RES-JB
    11. 8273
    Prev
    Next
    INDEX
    Explanations

    video footage from surveillance or security cameras

    oai_token-act-pair · gpt-3.5-turbo

    references to surveillance and video recordings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    olicy
    -0.69
    BuyableInstoreAndOnline
    -0.65
    ccoli
    -0.65
    ndra
    -0.64
    Else
    -0.64
    reddits
    -0.64
    isse
    -0.64
    erenn
    -0.64
    verning
    -0.63
    agnar
    -0.63
    POSITIVE LOGITS
     footage
    1.49
     recordings
    1.24
     tapes
    1.20
     surveillance
    1.19
     cameras
    1.17
     clips
    1.15
     filmed
    1.12
     tape
    1.12
     CCTV
    1.10
     evidence
    1.07
    Activations Density 0.102%

    No Known Activations