INDEX

Explanations

terms related to legal or diplomatic immunity

oai_token-act-pair · gpt-3.5-turbo

references to legal or protective immunity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

oric

-0.73

 Straw

-0.72

geist

-0.70

=-=-=-=-=-=-=-=-

-0.69

 Zeit

-0.67

Latest

-0.66

 Pens

-0.65

\\\\\\\\

-0.65

 Colleges

-0.64

product

-0.64

POSITIVE LOGITS

 immunity

1.73

 immune

0.95

 Immunity

0.90

 immun

0.80

imei

0.79

 defences

0.77

 pard

0.77

immune

0.77

 indemn

0.76

 comprom

0.75

Activations Density 0.005%

terms related to legal or diplomatic immunity

references to legal or protective immunity

No Comments

No Known Activations