Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 6-RES_POST_32K-OAI
    6. 4185
    Prev
    Next
    INDEX
    Explanations

    mentions of ghosts and related supernatural concepts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 6-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_6.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.6.hook_resid_post
    Hook Layer
    6
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    456
    +0.12
    0.5%
    571
    +0.11
    0.5%
    326
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.02
    571
    +0.11
    0.02
    549
    +0.10
    0.01
    Negative Logits
    zar
    -0.76
    utenberg
    -0.71
    erity
    -0.70
    baugh
    -0.68
    udeau
    -0.65
    Streamer
    -0.65
    dullah
    -0.65
    tics
    -0.65
    aid
    -0.64
    apsed
    -0.64
    POSITIVE LOGITS
     haunting
    0.96
    busters
    0.95
    writer
    0.95
    writing
    0.93
    writers
    0.89
     haunt
    0.89
    written
    0.87
     suspic
    0.77
    ghost
    0.76
    lore
    0.75
    Activations Density 0.079%

    No Known Activations