INDEX

Explanations

prepositions and their usage in the text

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

 constitutes

-0.66

QL

-0.63

catentry

-0.63

LV

-0.63

 represents

-0.62

 behaves

-0.61

ugal

-0.61

 publishes

-0.61

soType

-0.60

aces

-0.59

POSITIVE LOGITS

 jeopardy

0.80

escap

0.78

avering

0.76

 flux

0.72

versible

0.70

uble

0.69

noticed

0.66

putable

0.65

èĢ

0.65

 Wink

0.63

Activations Density 0.159%

prepositions and their usage in the text

No Comments

No Known Activations