INDEX

Explanations

expressions related to having expertise in a specific subject or field

oai_token-act-pair · gpt-3.5-turbo

references to expertise and specialized knowledge

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

gm

-0.64

ãĥİ

-0.64

 Shuttle

-0.63

miss

-0.61

mi

-0.60

early

-0.60

stant

-0.59

eve

-0.59

-0.58

van

-0.57

POSITIVE LOGITS

 expertise

0.97

 guiActiveUn

0.93

querque

0.83

anooga

0.83

ommel

0.82

afety

0.79

smanship

0.77

 credentials

0.75

linger

0.74

 chops

0.73

Activations Density 0.008%

expressions related to having expertise in a specific subject or field

references to expertise and specialized knowledge

No Comments

No Known Activations