INDEX

Explanations

the word "supreme" or related terms, as well as phrases related to authority and power

oai_token-act-pair · gpt-3.5-turbo

references to authority and supremacy

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

OUT

-0.90

ppo

-0.83

TPS

-0.74

okemon

-0.72

uffy

-0.70

ugg

-0.69

FORE

-0.69

ople

-0.67

zl

-0.66

kson

-0.65

POSITIVE LOGITS

ly

0.85

rament

0.80

most

0.76

essential

0.72

 secrecy

0.70

 vigilance

0.70

ITY

0.70

iour

0.69

doms

0.69

reme

0.68

Activations Density 0.013%

the word "supreme" or related terms, as well as phrases related to authority and power

references to authority and supremacy

No Comments

No Known Activations