INDEX

Explanations

dates in a specific format

oai_token-act-pair · gpt-3.5-turbo

dates and numerical references associated with events

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

avorite

-0.72

orable

-0.70

 secretaries

-0.67

ModLoader

-0.64

netflix

-0.63

imei

-0.62

 careless

-0.62

 habitual

-0.60

chwitz

-0.59

cific

-0.59

POSITIVE LOGITS

th

0.92

Interstitial

0.81

WH

0.76

0.75

Feb

0.73

0.72

TH

0.72

0.71

0.70

Activations Density 0.052%

dates in a specific format

dates and numerical references associated with events

No Comments

No Known Activations