© 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
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 7-RES-JB
    11. 21309
    Prev
    Next
    INDEX
    Explanations

    words related to concealing negative actions or traits, as well as social, ethical, and political issues

    oai_token-act-pair · gpt-3.5-turbo

    references to crime and legal issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    senal
    -0.65
     Vaugh
    -0.64
     Instr
    -0.63
    romeda
    -0.61
     helicop
    -0.59
     referen
    -0.59
     mosqu
    -0.59
     reluct
    -0.55
    querque
    -0.54
     encount
    -0.54
    POSITIVE LOGITS
    xia
    0.60
    )?
    0.59
    utterstock
    0.56
    '?
    0.52
    itarian
    0.51
    ability
    0.51
    Semitism
    0.51
    leness
    0.49
    iology
    0.48
    ulence
    0.48
    Activations Density 1.504%

    No Known Activations