INDEX

Explanations

technical terms and brand names related to technology

oai_token-act-pair · gpt-3.5-turbo

names of brands, companies, or notable entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ãĤ©

-0.89

RIS

-0.67

 caps

-0.65

bud

-0.64

 STEP

-0.63

oiler

-0.62

nap

-0.62

cknowled

-0.59

 FORM

-0.57

 shred

-0.57

POSITIVE LOGITS

elaide

0.97

endment

0.90

emis

0.86

qua

0.85

upid

0.84

doms

0.84

odore

0.82

psey

0.81

asonic

0.81

hyde

0.80

Activations Density 0.186%

technical terms and brand names related to technology

names of brands, companies, or notable entities

No Comments

No Known Activations