INDEX

Explanations

dates expressed numerically

oai_token-act-pair · gpt-3.5-turbo

numerical values associated with dates

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Genie

-0.68

igans

-0.62

igham

-0.60

ilater

-0.59

ukes

-0.58

urized

-0.58

aghetti

-0.57

ordable

-0.56

 secretaries

-0.56

ensical

-0.55

POSITIVE LOGITS

th

1.26

rd

0.98

0.90

TH

0.85

0.84

0.78

0.76

ths

0.76

0.75

Activations Density 0.077%

dates expressed numerically

numerical values associated with dates

No Comments

No Known Activations