INDEX

Explanations

instances of specific characters or letters occurring multiple times in the text

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

 partName

-0.74

ãĤĴ

-0.69

çĶ°

-0.66

ANCE

-0.65

OPLE

-0.63

 Carbuncle

-0.62

 Tsukuyomi

-0.61

 bluff

-0.61

 stuffing

-0.60

 Alleg

-0.59

POSITIVE LOGITS

acon

0.83

erman

0.82

ern

0.80

emonic

0.80

bit

0.75

bits

0.75

urious

0.71

abuse

0.70

carb

0.70

chel

0.69

Activations Density 0.020%

instances of specific characters or letters occurring multiple times in the text

No Comments

No Known Activations