INDEX

Explanations

words related to permission or ability to perform an action

oai_token-act-pair · gpt-3.5-turbo

references to the concept of capability or allowance in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Notting

-0.63

 Lump

-0.62

ravis

-0.61

gall

-0.60

govtrack

-0.59

ĪĴ

-0.58

çī

-0.58

_-

-0.58

ãĥĻ

-0.57

Sax

-0.57

POSITIVE LOGITS

 unrestricted

0.99

 flexibility

0.92

 unlimited

0.87

 access

0.86

 freely

0.85

 deduction

0.79

riott

0.75

 exceptions

0.72

 limitless

0.71

parency

0.71

Activations Density 0.193%

words related to permission or ability to perform an action

references to the concept of capability or allowance in various contexts

No Comments

No Known Activations