INDEX

Explanations

references to superlatives or extremes

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate the highest levels or amounts of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

izon

-0.70

redo

-0.70

vous

-0.68

icer

-0.65

ca

-0.65

md

-0.65

agra

-0.65

opard

-0.64

TeX

-0.64

chini

-0.63

POSITIVE LOGITS

 bidder

0.98

 rated

0.89

 imaginable

0.83

 pinnacle

0.82

 concentrations

0.82

 possible

0.81

 earners

0.81

 extent

0.78

 conceivable

0.77

 grade

0.77

Activations Density 0.022%

references to superlatives or extremes

phrases that indicate the highest levels or amounts of something

No Comments

No Known Activations