INDEX

Explanations

locations or positions of individuals in an image

oai_token-act-pair · gpt-3.5-turbo

directional indicators related to positioning in a visual context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

humans

-0.69

©¶æ¥µ

-0.67

-0.63

 Clicker

-0.61

universal

-0.61

Rule

-0.59

DERR

-0.58

Regular

-0.57

 Traps

-0.57

Lev

-0.56

POSITIVE LOGITS

 pose

0.91

 poses

0.84

 flanked

0.84

 screenshot

0.83

 caption

0.83

 photograph

0.82

 pictured

0.82

 posing

0.80

 photo

0.78

yright

0.78

Activations Density 0.047%

locations or positions of individuals in an image

directional indicators related to positioning in a visual context

No Comments

No Known Activations