INDEX

Explanations

phrases containing the word "true"

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

Pages

-0.75

uled

-0.74

acco

-0.71

ota

-0.70

hens

-0.69

adish

-0.69

rador

-0.68

cheon

-0.68

aid

-0.67

ovan

-0.67

POSITIVE LOGITS

 believers

1.02

 believer

0.99

 positives

0.85

 guiIcon

0.76

terday

0.75

çĭ

0.74

stic

0.74

 unfocusedRange

0.73

hood

0.69

 Detective

0.69

Activations Density 0.413%

phrases containing the word "true"

No Comments

No Known Activations