INDEX

Explanations

references to news events and media coverage

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

beh

-0.82

soDeliveryDate

-0.78

interstitial

-0.75

soType

-0.72

cffffcc

-0.72

genre

-0.70

 psychiat

-0.70

âĹ¼

-0.69

intent

-0.67

è£

-0.66

POSITIVE LOGITS

]'

0.85

ews

0.78

 Indy

0.77

:'

0.76

 Defeat

0.72

 Highlights

0.72

 Talks

0.71

 Differences

0.70

 McAuliffe

0.69

 Latest

0.69

Activations Density 0.059%

references to news events and media coverage

No Comments

No Known Activations