INDEX

Explanations

percentages or statistical figures

oai_token-act-pair · gpt-3.5-turbo

phrases that express statistical percentages or proportions

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

pedia

-0.80

ollah

-0.74

edin

-0.70

 caching

-0.68

ebin

-0.66

unct

-0.64

idi

-0.63

hran

-0.62

alph

-0.62

dl

-0.62

POSITIVE LOGITS

 respondents

0.89

pees

0.82

 households

0.81

GDP

0.80

 surveyed

0.73

 applicants

0.71

 voters

0.66

 variance

0.66

 Fortune

0.66

âķĲ

0.66

Activations Density 0.071%

percentages or statistical figures

phrases that express statistical percentages or proportions

No Comments

No Known Activations