INDEX

Explanations

terms related to reviewing, evaluating, or inspecting

oai_token-act-pair · gpt-3.5-turbo

terms related to veterinary and administrative contexts

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

theless

-0.99

lihood

-0.77

 Halls

-0.76

 Quarterly

-0.76

deen

-0.71

hani

-0.68

 unexplained

-0.67

 charism

-0.67

 Mond

-0.65

Lis

-0.65

POSITIVE LOGITS

hered

1.03

ocular

1.01

ours

0.94

onom

0.87

arious

0.85

nect

0.84

ulously

0.83

imated

0.81

bot

0.81

ument

0.81

Activations Density 0.040%

terms related to reviewing, evaluating, or inspecting

terms related to veterinary and administrative contexts

No Comments

No Known Activations