INDEX

Explanations

length-related terms or references within 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

Pwr

-0.77

 Camel

-0.73

eon

-0.69

 DRAG

-0.68

 Debor

-0.65

CV

-0.62

 laun

-0.62

 DPRK

-0.61

 upside

-0.61

 Shutterstock

-0.61

POSITIVE LOGITS

ovo

1.20

emies

1.10

omore

1.09

emy

0.96

meyer

0.95

quist

0.92

cious

0.91

vironment

0.91

opoly

0.90

esis

0.88

Activations Density 0.003%

length-related terms or references within the text

No Comments

No Known Activations