INDEX

Explanations

quantitative values

oai_token-act-pair · gpt-3.5-turbo

phrases and questions that express curiosity or inquiry

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

 Clarkson

-0.70

 Borders

-0.66

 Grail

-0.66

 Oval

-0.62

 Nare

-0.61

 submar

-0.61

 Berk

-0.61

 Gibraltar

-0.61

 Azerb

-0.60

edIn

-0.60

POSITIVE LOGITS

][

0.89

redients

0.76

voice

0.73

0.72

gpu

0.72

ql

0.70

0.69

))

0.69

vm

0.68

Activations Density 0.211%

quantitative values

phrases and questions that express curiosity or inquiry

No Comments

No Known Activations