INDEX

Explanations

situations or items that are beneficial or advantageous

oai_token-act-pair · gpt-3.5-turbo

phrases relating to the benefits or positive effects of various subjects

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

cles

-0.67

 Accessed

-0.65

REDACTED

-0.63

chenko

-0.63

Stan

-0.62

rose

-0.62

allah

-0.59

 traced

-0.59

stan

-0.59

urred

-0.59

POSITIVE LOGITS

geries

0.96

 example

0.93

 awhile

0.89

gotten

0.88

 purposes

0.84

gery

0.81

agers

0.80

aging

0.79

bidden

0.79

 detecting

0.78

Activations Density 0.125%

situations or items that are beneficial or advantageous

phrases relating to the benefits or positive effects of various subjects

No Comments

No Known Activations