INDEX

Explanations

phrases related to legal or governmental contexts

oai_token-act-pair · gpt-3.5-turbo

phrases related to quantities or measurements in specific contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

akening

-0.78

olls

-0.74

earing

-0.72

iled

-0.71

ernels

-0.70

-------------

-0.69

axter

-0.69

ivating

-0.66

ickle

-0.66

iling

-0.65

POSITIVE LOGITS

 conflic

0.89

 lang

0.78

 Organisation

0.77

dat

0.76

nu

0.75

Nou

0.70

 rÃ©

0.70

 physique

0.69

tem

0.67

Moy

0.67

Activations Density 0.074%

phrases related to legal or governmental contexts

phrases related to quantities or measurements in specific contexts

No Comments

No Known Activations