INDEX

Explanations

words related to years, dates, and time periods

oai_token-act-pair · gpt-3.5-turbo

significant years and related data points

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

CHAT

-0.60

plain

-0.56

¬¼

-0.55

acting

-0.55

 interfering

-0.55

Alternatively

-0.54

 guarding

-0.54

oyal

-0.54

anooga

-0.54

 sympath

-0.54

POSITIVE LOGITS

 Year

0.86

Year

0.83

0.82

0.78

0.75

0.74

 calendar

0.74

 YEAR

0.73

Activations Density 0.401%

words related to years, dates, and time periods

significant years and related data points

No Comments

No Known Activations