INDEX

Explanations

points of moral or ethical conflict and questioning boundaries or principles

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_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

DAQ

-0.84

CVE

-0.78

issance

-0.73

VERTISEMENT

-0.71

 unden

-0.70

âĺħâĺħ

-0.67

NECT

-0.67

=-=-=-=-=-=-=-=-

-0.66

oir

-0.66

sylvania

-0.66

POSITIVE LOGITS

 separating

1.11

breaking

1.00

 dividing

0.97

 between

0.95

 crossing

0.92

aments

0.90

breakers

0.87

 crossed

0.86

 crossings

0.82

posts

0.79

Activations Density 0.029%

points of moral or ethical conflict and questioning boundaries or principles

No Comments

No Known Activations