INDEX

Explanations

references to monetary amounts expressed as yearly figures

oai_token-act-pair · gpt-3.5-turbo

references to the term 'year' and its related metrics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

achus

-0.77

agonist

-0.77

ogle

-0.73

ortium

-0.72

 encl

-0.71

hod

-0.71

psey

-0.70

ktop

-0.69

 contag

-0.69

ocker

-0.67

POSITIVE LOGITS

ago

1.07

long

1.07

Ago

1.04

 olds

1.00

ns

0.88

glass

0.86

lene

0.84

lings

0.78

nings

0.76

ning

0.74

Activations Density 0.128%

references to monetary amounts expressed as yearly figures

references to the term 'year' and its related metrics

No Comments

No Known Activations