INDEX

Explanations

phrases that indicate a need or a call to action, particularly addressing the reader directly

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

 suffice

-0.72

forth

-0.66

 Faust

-0.64

bender

-0.63

 Hyde

-0.63

 Standing

-0.60

achable

-0.60

 Justice

-0.59

ÂłÂłÂłÂł

-0.58

tein

-0.57

POSITIVE LOGITS

've

0.98

isode

0.82

're

0.81

'll

0.79

 crave

0.77

 liked

0.69

orthy

0.69

 digest

0.68

 want

0.68

 favourite

0.68

Activations Density 0.013%

phrases that indicate a need or a call to action, particularly addressing the reader directly

No Comments

No Known Activations