INDEX

Explanations

dates written in a specific format like "Month Day" as in "June 1st"

oai_token-act-pair · gpt-3.5-turbo

dates, specifically the first day of a month

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

 extras

-0.55

 acron

-0.55

ioch

-0.55

iT

-0.54

natureconservancy

-0.53

KE

-0.52

 foreskin

-0.52

 laun

-0.52

 helicop

-0.52

 explan

-0.52

POSITIVE LOGITS

st

1.58

nd

1.46

rd

1.33

sta

1.01

ND

0.99

RD

0.96

nda

0.94

stal

0.92

stice

0.84

0.80

Activations Density 0.052%

dates written in a specific format like "Month Day" as in "June 1st"

dates, specifically the first day of a month

No Comments

No Known Activations