INDEX

Explanations

phrases starting with symbols like colons and greater than signs

oai_token-act-pair · gpt-3.5-turbo

colons and their associated content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ensibly

-0.80

senal

-0.76

 hemor

-0.70

 predec

-0.63

 temples

-0.60

 SEAL

-0.60

ngth

-0.60

 conflic

-0.60

 unden

-0.59

 notor

-0.59

POSITIVE LOGITS

 pige

0.74

Ð

0.64

"""

0.63

à¼

0.61

Dear

0.61

 Pist

0.61

rf

0.61

Requirements

0.60

 coli

0.60

Thread

0.60

Activations Density 0.065%

phrases starting with symbols like colons and greater than signs

colons and their associated content

No Comments

No Known Activations