INDEX

Explanations

Associated Press articles

oai_token-act-pair · gpt-3.5-turbo

references to the Associated Press

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

ysis

-0.95

ggle

-0.79

ppe

-0.79

isan

-0.75

wagen

-0.73

vette

-0.71

ppo

-0.69

jri

-0.69

ilk

-0.69

assi

-0.69

POSITIVE LOGITS

Images

0.90

 Images

0.78

 Newsp

0.77

notation

0.70

 associations

0.69

Journal

0.68

 destinations

0.67

iance

0.67

 newsp

0.67

 attractions

0.67

Activations Density 0.009%

Associated Press articles

references to the Associated Press

No Comments

No Known Activations