INDEX

Explanations

phrases related to unexpected developments or surprises

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

ufact

-0.73

ogl

-0.68

pta

-0.65

leased

-0.64

inez

-0.64

 Ducks

-0.64

 Domain

-0.63

isance

-0.63

league

-0.63

rahim

-0.62

POSITIVE LOGITS

 twist

1.25

 twists

1.16

 Twist

0.87

 twisting

0.85

 weave

0.80

 Whedon

0.80

 endings

0.77

 knot

0.73

hered

0.72

 rope

0.71

Activations Density 0.022%

phrases related to unexpected developments or surprises

No Comments

No Known Activations