INDEX

Explanations

references to real-life scenarios and events

oai_token-act-pair · gpt-3.5-turbo

references to "real life" concepts or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Hyper

-0.72

 shoulders

-0.69

uned

-0.67

 rounding

-0.66

MENTS

-0.66

IAL

-0.65

 hyper

-0.65

sonian

-0.65

 Cosponsors

-0.64

 Burk

-0.64

POSITIVE LOGITS

life

1.14

guards

0.94

 expectancy

0.92

 experien

0.92

lihood

0.91

mares

0.89

guard

0.84

theless

0.82

lif

0.79

hazard

0.76

Activations Density 0.017%

references to real-life scenarios and events

references to "real life" concepts or situations

No Comments

No Known Activations