INDEX

Explanations

words related to personal qualities or attributes

oai_token-act-pair · gpt-3.5-turbo

concepts related to value, worth, and potential

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

NSA

-0.72

Vs

-0.70

wered

-0.69

requently

-0.68

sg

-0.68

otom

-0.66

articles

-0.64

RIC

-0.64

LB

-0.64

Ban

-0.63

POSITIVE LOGITS

 stature

0.89

smanship

0.82

 abilities

0.79

 sensibilities

0.78

 ability

0.75

 esteem

0.74

iness

0.74

 reputation

0.73

 posture

0.73

lust

0.73

Activations Density 0.386%

words related to personal qualities or attributes

concepts related to value, worth, and potential

No Comments

No Known Activations