INDEX

Explanations

phrases or words related to basic concepts or principles

oai_token-act-pair · gpt-3.5-turbo

references to fundamental concepts or necessities

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

romeda

-0.79

igham

-0.74

oping

-0.73

ibal

-0.70

 Tycoon

-0.70

isSpecialOrderable

-0.67

encer

-0.67

 Rodrigo

-0.66

 prow

-0.65

estern

-0.64

POSITIVE LOGITS

 necessities

1.10

 tenets

1.04

basic

0.98

 arithmetic

0.92

lly

0.89

 principles

0.88

 premise

0.81

gradient

0.80

 outline

0.79

essential

0.78

Activations Density 0.022%

phrases or words related to basic concepts or principles

references to fundamental concepts or necessities

No Comments

No Known Activations