INDEX

Explanations

phrases indicating the non-existence of something that is commonly believed or expected

oai_token-act-pair · gpt-3.5-turbo

phrases that suggest the non-existence of certain concepts or entities

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

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

rik

-0.81

hend

-0.74

alde

-0.72

uay

-0.72

 Heard

-0.67

WARD

-0.66

Roberts

-0.65

ÄŁ

-0.65

qqa

-0.63

OIL

-0.63

POSITIVE LOGITS

 thing

1.66

 entity

1.11

thing

1.09

 Thing

0.99

 person

0.90

 creature

0.90

 phenomenon

0.86

 instance

0.86

THING

0.85

 facility

0.83

Activations Density 0.041%

phrases indicating the non-existence of something that is commonly believed or expected

phrases that suggest the non-existence of certain concepts or entities

No Comments

No Known Activations