INDEX

Explanations

words related to negative or alarming events or situations

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

á

-0.85

oun

-0.81

elta

-0.79

agos

-0.77

uay

-0.74

guided

-0.73

onso

-0.73

ardi

-0.73

glas

-0.72

pei

-0.71

POSITIVE LOGITS

ly

1.02

ingly

0.98

 revelations

0.84

 Revelations

0.83

 Trend

0.82

Trend

0.80

 behaviour

0.80

LY

0.79

 trends

0.78

 incidents

0.77

Activations Density 0.054%

words related to negative or alarming events or situations

No Comments

No Known Activations