INDEX

Explanations

phrases related to technology and organization communication

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

 ment

-0.65

nesota

-0.64

 destro

-0.63

 cannabin

-0.61

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.61

ï¸

-0.60

ħĭ

-0.60

inventoryQuantity

-0.60

 condem

-0.59

lax

-0.59

POSITIVE LOGITS

ribute

1.42

ractor

1.32

ARGET

1.32

olerance

1.30

ributes

1.29

ribut

1.24

ension

1.23

ensions

1.21

itled

1.20

itles

1.18

Activations Density 1.080%

phrases related to technology and organization communication

No Comments

No Known Activations