INDEX

Explanations

references to specific physical locations or positions within a space

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

cci

-0.82

ensable

-0.79

anned

-0.76

rss

-0.76

ardy

-0.76

wana

-0.74

 Parenthood

-0.71

endi

-0.71

atoon

-0.70

atis

-0.70

POSITIVE LOGITS

side

1.14

most

1.07

 lobe

0.96

 side

0.95

 thereof

0.93

wards

0.90

 facing

0.87

 porch

0.86

ward

0.86

facade

0.83

Activations Density 1.524%

references to specific physical locations or positions within a space

No Comments

No Known Activations