INDEX

Explanations

statistical likelihoods or probabilities

oai_token-act-pair · gpt-3.5-turbo

phrases conveying likelihood or probability

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

limits

-0.79

Cooldown

-0.73

comings

-0.71

books

-0.71

Graphics

-0.69

Wra

-0.69

 pandemonium

-0.67

holder

-0.67

cedented

-0.67

ä¹ĭ

-0.66

POSITIVE LOGITS

 succeed

1.15

be

1.10

 cite

1.07

 offend

1.07

 conclude

1.06

 participate

1.05

 receive

1.05

 survive

1.04

 perceive

1.02

 engage

1.01

Activations Density 0.085%

statistical likelihoods or probabilities

phrases conveying likelihood or probability

No Comments

No Known Activations