INDEX

Explanations

references to research studies and observations in the form "studies have shown"

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "have," particularly in contexts discussing studies or findings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

oppable

-0.67

 Refresh

-0.66

Thank

-0.63

 Merchant

-0.63

ickle

-0.63

 fulfil

-0.62

inance

-0.62

 Thank

-0.60

ttp

-0.60

 inflic

-0.60

POSITIVE LOGITS

 been

1.29

 begun

1.12

 speculated

1.09

 argued

1.07

 shown

1.07

 warned

1.06

 suggested

1.05

 debunked

1.05

 gotten

1.03

 theor

1.01

Activations Density 0.206%

references to research studies and observations in the form "studies have shown"

phrases that include the word "have," particularly in contexts discussing studies or findings

No Comments

No Known Activations