INDEX

Explanations

terms related to authoritative figures or actions of authority

oai_token-act-pair · gpt-3.5-turbo

phrases related to criminal activity or legal issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

irk

-0.67

hra

-0.64

rift

-0.63

itsch

-0.62

epend

-0.61

UV

-0.61

ello

-0.59

çĦ

-0.58

ulet

-0.58

azard

-0.57

POSITIVE LOGITS

 someday

1.14

 tomorrow

1.03

 freaking

0.93

 anytime

0.91

 goddamn

0.89

ANY

0.87

illions

0.87

 idiots

0.85

 forever

0.83

 EVERY

0.82

Activations Density 1.155%

terms related to authoritative figures or actions of authority

phrases related to criminal activity or legal issues

No Comments

No Known Activations