INDEX

Explanations

phrases related to legal issues and fines

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soDeliveryDate

-0.87

ern

-0.73

atl

-0.73

 awaits

-0.71

meric

-0.70

wang

-0.69

NET

-0.67

GROUP

-0.66

ise

-0.65

 Mehran

-0.65

POSITIVE LOGITS

 violating

1.05

 mishand

1.03

 daring

1.00

 misconduct

0.96

 transgress

0.94

 tresp

0.94

 negligence

0.93

 breaching

0.93

 failing

0.93

 inaction

0.92

Activations Density 2.567%

phrases related to legal issues and fines

No Comments

No Known Activations