INDEX

Explanations

references to locations or specific geographical areas

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

kefeller

-0.75

iosyncr

-0.72

 hindsight

-0.66

eas

-0.66

enegger

-0.66

catentry

-0.66

 forgiven

-0.65

phony

-0.64

ENTION

-0.63

 Bild

-0.62

POSITIVE LOGITS

gow

0.86

 Lumpur

0.85

eston

0.82

Clark

0.77

 Jenner

0.69

thia

0.68

ath

0.68

auga

0.67

à

0.66

zos

0.66

Activations Density 0.025%

references to locations or specific geographical areas

No Comments

No Known Activations