INDEX

Explanations

medical and safety-related terms and procedures often related to accidents, first aid, and health

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

creen

-0.97

rooms

-0.89

leaders

-0.87

geon

-0.85

ching

-0.82

eworthy

-0.81

kie

-0.81

sonian

-0.80

wagen

-0.79

roads

-0.79

POSITIVE LOGITS

uates

0.71

eering

0.69

 gratification

0.69

ALLY

0.68

mble

0.66

ments

0.66

IBLE

0.64

 replay

0.64

ensity

0.62

aneous

0.61

Activations Density 4.926%

medical and safety-related terms and procedures often related to accidents, first aid, and health

No Comments

No Known Activations