INDEX

Explanations

terms related to stealth and secrecy

oai_token-act-pair · gpt-3.5-turbo

terms related to stealth and health

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

Interstitial

-0.68

ocese

-0.68

ONT

-0.67

artney

-0.66

ICAN

-0.66

Pwr

-0.66

onz

-0.65

ãĥ¤

-0.61

ktop

-0.60

thritis

-0.60

POSITIVE LOGITS

ily

1.09

ibility

1.02

ively

0.94

iest

0.94

iness

0.94

 bomber

0.87

ibly

0.86

 stealth

0.85

 camouflage

0.85

loo

0.83

Activations Density 0.014%

terms related to stealth and secrecy

terms related to stealth and health

No Comments

No Known Activations