INDEX

Explanations

phrases related to components or aspects of a larger entity or concept

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "part."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ãĥ¼ãĥ³

-0.60

glers

-0.59

unders

-0.59

çĦ

-0.58

âĤ¬

-0.56

æ©

-0.55

balance

-0.55

 pores

-0.53

opa

-0.53

ges

-0.53

POSITIVE LOGITS

icularly

1.10

aking

1.05

ially

0.99

nered

0.96

icular

0.95

of

0.79

heartedly

0.77

ridges

0.75

hearted

0.74

isan

0.73

Activations Density 0.035%

phrases related to components or aspects of a larger entity or concept

phrases that include the word "part."

No Comments

No Known Activations