INDEX

Explanations

phrases related to being at the top or high ranking positions

oai_token-act-pair · gpt-3.5-turbo

phrases involving "top of" in relation to rankings or positions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iors

-0.70

ary

-0.66

mere

-0.65

 survives

-0.64

Cause

-0.64

fy

-0.64

rosso

-0.64

spring

-0.63

etta

-0.62

PsyNetMessage

-0.62

POSITIVE LOGITS

 heap

1.02

 hill

0.99

 stairs

0.98

 pyramid

0.89

 totem

0.87

 charts

0.85

 pile

0.85

 earners

0.85

 notch

0.83

 ladder

0.82

Activations Density 0.078%

phrases related to being at the top or high ranking positions

phrases involving "top of" in relation to rankings or positions

No Comments

No Known Activations