INDEX

Explanations

desires or intentions expressed through the word "want."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asse

-0.71

rir

-0.65

icist

-0.65

frac

-0.64

NVIDIA

-0.62

icol

-0.61

ashing

-0.60

iverpool

-0.59

ibliography

-0.58

scar

-0.57

POSITIVE LOGITS

reprene

0.96

 everybody

0.86

 everyone

0.85

to

0.81

 answers

0.80

 somebody

0.78

 someone

0.77

 revenge

0.76

 clarification

0.73

 people

0.73

Activations Density 0.082%

desires or intentions expressed through the word "want."

No Comments

No Known Activations