INDEX

Explanations

phrases related to making or causing something to happen

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

yd

-0.67

 consulted

-0.66

bow

-0.61

Ha

-0.60

ban

-0.58

\-

-0.58

ologue

-0.56

 signed

-0.56

 modeled

-0.56

 tweeted

-0.56

POSITIVE LOGITS

us

1.01

him

0.80

 them

0.77

me

0.76

 tremend

0.76

SPONSORED

0.73

 viewers

0.71

olves

0.71

 havoc

0.70

 investors

0.70

Activations Density 1.416%

phrases related to making or causing something to happen

No Comments

No Known Activations