INDEX

Explanations

time-related phrases indicating durations or lengths, usually involving years, weeks, or hours

oai_token-act-pair · gpt-3.5-turbo

references to durations of time and the concept of years

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ooter

-0.72

 exception

-0.67

llor

-0.66

ocard

-0.65

apple

-0.63

ERAL

-0.62

agame

-0.61

sheet

-0.61

atche

-0.61

XY

-0.60

POSITIVE LOGITS

chool

0.91

entimes

0.86

ago

0.82

 afterward

0.80

 upon

0.78

 afterwards

0.74

aults

0.74

ayers

0.73

ourcing

0.71

 past

0.71

Activations Density 0.071%

time-related phrases indicating durations or lengths, usually involving years, weeks, or hours

references to durations of time and the concept of years

No Comments

No Known Activations