INDEX

Explanations

contractions involving "don't" and the word "don't" by itself

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "don't" or its variations

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

 Reality

-0.66

 Butt

-0.64

 Remastered

-0.63

cised

-0.62

 Completed

-0.61

 DRAGON

-0.61

 Advice

-0.61

 afore

-0.60

 Dise

-0.60

 Personality

-0.60

POSITIVE LOGITS

't

1.35

ned

0.99

ning

0.85

ates

0.82

ners

0.80

ated

0.78

etsk

0.78

nels

0.74

ate

0.74

uts

0.74

Activations Density 0.106%

contractions involving "don't" and the word "don't" by itself

instances of the phrase "don't" or its variations

No Comments

No Known Activations