INDEX

Explanations

specific mentions of lectures or lecture-related activities

oai_token-act-pair · gpt-3.5-turbo

references to lectures

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

anking

-0.70

odka

-0.69

adjusted

-0.66

uilt

-0.66

bably

-0.65

ansas

-0.56

 Cherry

-0.55

 preservation

-0.55

 wild

-0.55

 Belarus

-0.54

POSITIVE LOGITS

 Lect

1.13

ures

0.92

gow

0.89

shire

0.87

ysis

0.81

hus

0.81

inar

0.81

iple

0.79

¿½

0.79

itle

0.79

Activations Density 0.014%

specific mentions of lectures or lecture-related activities

references to lectures

No Comments

No Known Activations