INDEX

Explanations

fraction or proportion phrases related to statistics

oai_token-act-pair · gpt-3.5-turbo

phrases expressing statistical ratios or proportions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

complex

-0.72

ande

-0.72

 ramp

-0.69

gang

-0.66

Stage

-0.65

sort

-0.64

rev

-0.64

heng

-0.64

wcsstore

-0.62

andan

-0.60

POSITIVE LOGITS

 Americans

0.93

 respondents

0.79

 households

0.77

(%

0.74

 American

0.73

 earners

0.73

 adults

0.71

 Australians

0.71

 kcal

0.70

ones

0.70

Activations Density 0.087%

fraction or proportion phrases related to statistics

phrases expressing statistical ratios or proportions

No Comments

No Known Activations