INDEX

Explanations

instances of violence or conflict-related terms and descriptions

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

hend

-0.78

ï¸

-0.77

 Boone

-0.75

Õ

-0.72

ysics

-0.72

uay

-0.69

ruary

-0.69

arden

-0.68

 Heard

-0.68

rition

-0.67

POSITIVE LOGITS

 thing

0.92

 instance

0.89

 cases

0.79

 contra

0.77

 example

0.74

 device

0.73

 facility

0.72

 garment

0.71

 defect

0.70

 individual

0.70

Activations Density 0.021%

instances of violence or conflict-related terms and descriptions

No Comments

No Known Activations