INDEX

Explanations

stand<stand> actions or situations where someone stands or has the potential to stand up for something, whether physically or metaphorically

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

allo

-0.75

erm

-0.67

javascript

-0.65

iler

-0.65

ilies

-0.64

iling

-0.63

 CLSID

-0.62

ilic

-0.61

selage

-0.61

eco

-0.59

POSITIVE LOGITS

still

1.08

 atop

1.01

 upright

0.96

 seiz

0.89

 silently

0.85

 idle

0.84

 patiently

0.84

 beside

0.83

iate

0.81

 tall

0.80

Activations Density 1.395%

stand<stand> actions or situations where someone stands or has the potential to stand up for something, whether physically or metaphorically

No Comments

No Known Activations