INDEX

Explanations

explanations or descriptions of how something works

oai_token-act-pair · gpt-3.5-turbo

phrases and expressions related to explaining how something functions or operates

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

imar

-0.80

miss

-0.70

hawks

-0.69

osi

-0.68

azines

-0.68

ulp

-0.66

mitt

-0.65

ĳå£«

-0.64

vette

-0.64

azar

-0.64

POSITIVE LOGITS

eth

0.81

 Myster

0.69

:[

0.67

 proced

0.65

 structured

0.65

places

0.64

 Anyway

0.64

 unravel

0.62

 disse

0.62

place

0.61

Activations Density 0.086%

explanations or descriptions of how something works

phrases and expressions related to explaining how something functions or operates

No Comments

No Known Activations