INDEX

Explanations

numerical values appearing in a specific format

oai_token-act-pair · gpt-3.5-turbo

references to specific numerical values or identifiers

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

enzie

-0.91

orney

-0.81

aeus

-0.81

ype

-0.80

yle

-0.79

aleb

-0.77

achus

-0.77

achu

-0.77

anooga

-0.75

icles

-0.74

POSITIVE LOGITS

th

0.98

0.86

0.82

nd

0.80

0.79

0.76

mph

0.74

teen

0.73

0.71

0.70

Activations Density 0.025%

numerical values appearing in a specific format

references to specific numerical values or identifiers

No Comments

No Known Activations