INDEX

Explanations

instances of phrases suggesting comparison or emphasis

oai_token-act-pair · gpt-3.5-turbo

declarative phrases that introduce or condition statements

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

Ruk

-0.72

 Divinity

-0.65

 Doctrine

-0.64

hips

-0.63

Eye

-0.62

istics

-0.61

Dog

-0.61

 Bull

-0.59

 Gallery

-0.59

Mis

-0.58

POSITIVE LOGITS

 surrounds

0.85

 consumes

0.84

 accompanies

0.84

 encompasses

0.83

 carbohyd

0.82

 extends

0.82

¥µ

0.82

ÃĥÃĤ

0.82

 occurs

0.81

 arose

0.81

Activations Density 0.229%

instances of phrases suggesting comparison or emphasis

declarative phrases that introduce or condition statements

No Comments

No Known Activations