INDEX

Explanations

references to awards and nominations

oai_token-act-pair · gpt-3.5-turbo

accolades and awards related to achievements in arts and entertainment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 squats

-0.81

 hops

-0.66

 dele

-0.66

 pads

-0.65

rone

-0.64

 decomp

-0.63

aucuses

-0.63

 practise

-0.63

 wakes

-0.63

 flare

-0.62

POSITIVE LOGITS

 Best

1.31

 Critics

1.27

Best

1.17

 Cannes

1.15

 Excellence

1.14

 Editors

1.10

 Awards

1.06

 accol

1.06

 awards

1.05

 Jury

1.04

Activations Density 0.259%

references to awards and nominations

accolades and awards related to achievements in arts and entertainment

No Comments

No Known Activations