INDEX

Explanations

terms related to specific car models

oai_token-act-pair · gpt-3.5-turbo

mentions of specific academic or institutional names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

EFF

-0.71

 APPLIC

-0.68

 accomp

-0.67

 artif

-0.65

 enthusi

-0.62

 begging

-0.61

 cousin

-0.61

 epis

-0.61

 misdem

-0.60

 framing

-0.59

POSITIVE LOGITS

isoft

0.93

ampire

0.91

endish

0.89

idable

0.88

netic

0.85

yang

0.85

ixt

0.84

raid

0.82

letal

0.80

auld

0.80

Activations Density 0.187%

terms related to specific car models

mentions of specific academic or institutional names

No Comments

No Known Activations