INDEX

Explanations

time spans or durations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "to" and its varied contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

erto

-0.78

creat

-0.71

soDeliveryDate

-0.71

 stru

-0.67

 retri

-0.66

intosh

-0.64

RF

-0.64

irez

-0.63

need

-0.63

 vulner

-0.61

POSITIVE LOGITS

pload

0.90

ggles

0.90

 ensure

0.89

 fulfill

0.83

 shore

0.82

asted

0.81

 satisfy

0.81

 eliminate

0.80

 achieve

0.80

 compensate

0.79

Activations Density 0.313%

time spans or durations

occurrences of the word "to" and its varied contexts

No Comments

No Known Activations