INDEX

Explanations

words related to authenticity and authentication

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

storms

-0.71

ower

-0.63

GPU

-0.62

inson

-0.62

division

-0.61

poon

-0.61

sol

-0.60

 allocations

-0.60

owers

-0.59

esville

-0.59

POSITIVE LOGITS

 authenticity

0.95

 authentic

0.93

icator

0.91

 Authent

0.91

 authent

0.85

authent

0.79

imated

0.77

uine

0.77

icating

0.77

icas

0.76

Activations Density 0.029%

words related to authenticity and authentication

No Comments

No Known Activations