INDEX

Explanations

phrases expressing recommendation or advice

oai_token-act-pair · gpt-3.5-turbo

phrases that express recommendations or obligations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Patty

-0.61

CI

-0.61

Fra

-0.58

vati

-0.57

 syndrome

-0.56

stood

-0.54

Oss

-0.54

 Afgh

-0.54

assi

-0.54

noticed

-0.54

POSITIVE LOGITS

ered

1.23

ering

1.14

be

1.13

 ideally

1.08

 suffice

0.96

nt

0.95

erers

0.87

 theoretically

0.84

 strive

0.83

Activations Density 0.067%

phrases expressing recommendation or advice

phrases that express recommendations or obligations

No Comments

No Known Activations