INDEX

Explanations

the initials "SI" in various contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of the abbreviation "SI" and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

lessly

-0.73

killer

-0.70

score

-0.69

raising

-0.68

stage

-0.68

places

-0.68

chains

-0.67

actionDate

-0.67

rals

-0.66

spread

-0.66

POSITIVE LOGITS

pler

0.88

adoes

0.79

zona

0.73

avage

0.71

pling

0.71

UE

0.71

uous

0.70

OLOG

0.69

obar

0.68

 Appendix

0.68

Activations Density 0.036%

the initials "SI" in various contexts

mentions of the abbreviation "SI" and related terms

No Comments

No Known Activations