INDEX

Explanations

words related to quotes or praises that convey a strong positive sentiment

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

AGES

-0.72

strap

-0.69

croft

-0.67

HAEL

-0.67

 Lyme

-0.65

 revealing

-0.63

 testosterone

-0.63

 warming

-0.62

 Painter

-0.61

 wounding

-0.61

POSITIVE LOGITS

arantine

1.18

isition

1.17

ipped

1.09

idd

1.09

irrel

1.05

inoa

1.04

oted

1.04

arie

1.04

artz

1.03

isitions

1.01

Activations Density 0.531%

words related to quotes or praises that convey a strong positive sentiment

No Comments

No Known Activations