INDEX

Explanations

mentions of authority figures or supervisors

oai_token-act-pair · gpt-3.5-turbo

references to figures of authority or leadership, specifically the term "boss."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Scotia

-0.78

CTR

-0.66

 IMAGES

-0.61

Utah

-0.61

 Democr

-0.60

 dehyd

-0.60

 Trails

-0.58

RIPT

-0.58

 Featured

-0.58

åĤ

-0.57

POSITIVE LOGITS

hips

0.94

anova

0.91

enger

0.80

umo

0.80

ial

0.80

xon

0.79

boss

0.75

imil

0.75

imo

0.75

nuts

0.74

Activations Density 0.035%

mentions of authority figures or supervisors

references to figures of authority or leadership, specifically the term "boss."

No Comments

No Known Activations