INDEX

Explanations

references to the character "M" in various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Mechdragon

-0.69

 hottest

-0.66

 yours

-0.65

 skim

-0.65

 Mellon

-0.65

 Strip

-0.64

 Eleven

-0.63

Bil

-0.62

 blot

-0.62

 fashioned

-0.62

POSITIVE LOGITS

OST

1.19

ISSION

1.17

ixed

1.16

iscal

1.16

uzzle

1.16

otive

1.15

useum

1.15

ountain

1.15

umbai

1.12

igrant

1.12

Activations Density 0.034%

references to the character "M" in various contexts

No Comments

No Known Activations