INDEX

Explanations

interjections or expressions of surprise and emotion

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

ĪĴ

-0.90

ãĤ¼ãĤ¦ãĤ¹

-0.67

ãĥĩ

-0.67

 Edison

-0.65

 Metatron

-0.65

è¦ļéĨĴ

-0.64

uggest

-0.64

ãĥĥãĥī

-0.62

 Guant

-0.62

Pwr

-0.61

POSITIVE LOGITS

glers

0.90

oslav

0.87

ety

0.85

eworthy

0.82

atory

0.75

chers

0.74

hett

0.72

wear

0.72

kee

0.71

odon

0.70

Activations Density 0.009%

interjections or expressions of surprise and emotion

No Comments

No Known Activations