INDEX

Explanations

phrases related to inclusivity or the incorporation of various elements

oai_token-act-pair · gpt-3.5-turbo

references to various categories or examples within a text

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

mosp

-0.60

elling

-0.59

 Cummings

-0.59

 Lung

-0.58

 Oaks

-0.56

 Oakland

-0.56

icultural

-0.55

raq

-0.55

Preview

-0.54

 Vaughan

-0.54

POSITIVE LOGITS

itiz

0.76

 guiActiveUn

0.74

iton

0.71

hots

0.70

available

0.67

fman

0.65

ser

0.65

BF

0.64

atta

0.64

gradient

0.63

Activations Density 0.150%

phrases related to inclusivity or the incorporation of various elements

references to various categories or examples within a text

No Comments

No Known Activations