INDEX

Explanations

phrases related to refraining or abstaining from something

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of "usage" or "being used."

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

hill

-0.77

estial

-0.76

ontent

-0.74

izoph

-0.72

jri

-0.70

opic

-0.68

acters

-0.67

anche

-0.66

amac

-0.66

 corrid

-0.66

POSITIVE LOGITS

 Bauer

0.80

 Investigator

0.70

 innocence

0.69

 acquaintance

0.69

 excuses

0.68

Footnote

0.68

 oneself

0.67

 Prosecut

0.66

 dismiss

0.66

enary

0.65

Activations Density 0.045%

phrases related to refraining or abstaining from something

terms related to the concept of "usage" or "being used."

No Comments

No Known Activations