INDEX

Explanations

phrases related to guided tours or sightseeing

oai_token-act-pair · gpt-3.5-turbo

references to tours

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 procedural

-0.66

 penalties

-0.66

 penalty

-0.64

 suppressed

-0.63

 loose

-0.63

 hurry

-0.62

 predictive

-0.62

corn

-0.62

 forfeit

-0.62

 fluid

-0.61

POSITIVE LOGITS

 tour

1.04

neys

1.03

 tours

1.00

ney

0.97

 Tours

0.95

neau

0.90

icular

0.83

efully

0.81

 Tour

0.80

ogue

0.79

Activations Density 0.020%

phrases related to guided tours or sightseeing

references to tours

No Comments

No Known Activations