© 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. 17844
    Prev
    Next
    INDEX
    Explanations

    numbers related to a specific SCP entity

    oai_token-act-pair · gpt-3.5-turbo

    references to specific SCP (Secure, Contain, Protect) items and their classifications

    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
    ofi
    -0.77
    ornia
    -0.71
    onga
    -0.69
    amer
    -0.60
    Ô
    -0.60
     wi
    -0.60
    ITH
    -0.59
     dot
    -0.58
    anguages
    -0.58
     salute
    -0.58
    POSITIVE LOGITS
    ²¾
    0.84
    âĸĪâĸĪâĸĪâĸĪ
    0.80
    ãĥ¤
    0.80
    involved
    0.79
    âĸĪâĸĪ
    0.76
    XXXX
    0.73
    Ĥİ
    0.72
    ¶
    0.71
    Meta
    0.70
    ########
    0.70
    Activations Density 0.102%

    No Known Activations