INDEX

Explanations

titles or references to authoritative figures

oai_token-act-pair · gpt-3.5-turbo

references to authority figures and their roles, particularly those related to the term "lord" and "woman"

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

¥ŀ

-0.86

unes

-0.66

 skelet

-0.65

 Citiz

-0.65

 Palestin

-0.64

 widest

-0.61

itialized

-0.60

 lightweight

-0.60

 insulation

-0.59

ogi

-0.59

POSITIVE LOGITS

lord

0.97

hood

0.85

lords

0.81

hyde

0.80

pool

0.76

der

0.76

ëĭ

0.75

hattan

0.75

ifest

0.74

ipop

0.74

Activations Density 0.022%

titles or references to authoritative figures

references to authority figures and their roles, particularly those related to the term "lord" and "woman"

No Comments

No Known Activations