INDEX

Explanations

references to family members, especially uncles

oai_token-act-pair · gpt-3.5-turbo

mentions of familial relationships, specifically related to uncles

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

mberg

-0.95

ãĤ¤ãĥĪ

-0.69

 largeDownload

-0.67

istas

-0.65

 lapse

-0.64

EVA

-0.63

punk

-0.63

Effect

-0.62

NG

-0.62

*/(

-0.61

POSITIVE LOGITS

 uncle

1.14

liest

0.94

hest

0.92

 Uncle

0.88

hetically

0.86

 nephew

0.85

 aunt

0.83

heses

0.82

 niece

0.80

ired

0.79

Activations Density 0.011%

references to family members, especially uncles

mentions of familial relationships, specifically related to uncles

No Comments

No Known Activations