INDEX

Explanations

phrases related to analyzing or commenting on how someone or something appears, performs, or behaves

oai_token-act-pair · gpt-3.5-turbo

instances of the verb "look" in various forms

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

Osw

-0.77

âĹ¼

-0.75

orb

-0.74

SPONSORED

-0.70

 Loch

-0.67

icipated

-0.67

unes

-0.67

wealth

-0.66

stem

-0.65

venient

-0.60

POSITIVE LOGITS

 forward

0.76

 suspic

0.74

 poised

0.73

 favorably

0.72

ahead

0.70

 suspicious

0.69

 shif

0.68

ared

0.67

ENE

0.67

ãĤ¶

0.65

Activations Density 0.051%

phrases related to analyzing or commenting on how someone or something appears, performs, or behaves

instances of the verb "look" in various forms

No Comments

No Known Activations