INDEX

Explanations

phrases comparing or diminishing one thing to another

oai_token-act-pair · gpt-3.5-turbo

phrases that diminish the significance of something by suggesting it is "nothing more than" a trivial or lesser version of itself

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ahime

-0.83

ode

-0.80

anta

-0.77

enser

-0.70

NCT

-0.68

-0.65

oris

-0.65

arts

-0.65

oren

-0.63

rones

-0.62

POSITIVE LOGITS

 mediocre

0.70

 rudimentary

0.69

 filler

0.67

anke

0.66

 cosmetic

0.65

 subsistence

0.64

 superficial

0.62

 pure

0.61

 hors

0.60

0.59

Activations Density 0.059%

phrases comparing or diminishing one thing to another

phrases that diminish the significance of something by suggesting it is "nothing more than" a trivial or lesser version of itself

No Comments

No Known Activations