INDEX

Explanations

terms related to reading and writing data in programming contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

 Goodman

-0.71

Zur

-0.69

 Conquest

-0.69

Aus

-0.64

Strike

-0.63

asia

-0.63

anova

-0.62

giving

-0.60

 Invasion

-0.59

xon

-0.57

POSITIVE LOGITS

just

1.08

ying

1.05

ied

1.03

 aloud

0.95

dress

0.88

mitt

0.87

mitted

0.86

minist

0.83

dir

0.81

ahead

0.81

Activations Density 0.020%

terms related to reading and writing data in programming contexts

No Comments

No Known Activations