INDEX

Explanations

references to needing or granting permission

oai_token-act-pair · gpt-3.5-turbo

references to the concept of permission

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

ulz

-0.72

tal

-0.70

store

-0.67

 Tycoon

-0.65

ulative

-0.64

nik

-0.64

 Springfield

-0.64

chet

-0.63

 Chrom

-0.63

 Revel

-0.62

POSITIVE LOGITS

 permission

1.05

 confir

0.94

 permissions

0.91

Reviewer

0.90

ibl

0.82

essee

0.81

 slips

0.79

 granted

0.78

 waivers

0.77

mares

0.75

Activations Density 0.015%

references to needing or granting permission

references to the concept of permission

No Comments

No Known Activations