INDEX

Explanations

action verbs in the form of present tense or infinitive verbs

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.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

 curiously

-0.95

 wisely

-0.95

 oddly

-0.88

 cautiously

-0.88

 invariably

-0.86

 appropriately

-0.86

 obviously

-0.86

 optionally

-0.85

 generally

-0.85

 typically

-0.85

POSITIVE LOGITS

hing

0.71

rez

0.69

anism

0.67

ented

0.64

ifiable

0.63

yth

0.63

PsyNetMessage

0.62

ourced

0.59

Ñģ

0.59

ective

0.59

Activations Density 0.546%

action verbs in the form of present tense or infinitive verbs

No Comments

No Known Activations