INDEX

Explanations

references to inclusivity or collective terms

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

hran

-0.87

yip

-0.72

culosis

-0.69

Gw

-0.63

IND

-0.62

KH

-0.61

artz

-0.61

hz

-0.60

isu

-0.60

Nap

-0.59

POSITIVE LOGITS

usions

1.00

iances

0.98

uding

0.96

 attendant

0.94

udes

0.93

 kinds

0.91

 associated

0.84

 alike

0.84

iance

0.83

ocating

0.83

Activations Density 0.051%

references to inclusivity or collective terms

No Comments

No Known Activations