INDEX

Explanations

entities or items listed in a series

oai_token-act-pair · gpt-3.5-turbo

references to different groups, individuals, and discussions related to various topics

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

click

-0.71

reply

-0.65

 Peck

-0.62

opter

-0.61

okÃ©

-0.61

GPU

-0.61

 Nickel

-0.60

Corp

-0.60

better

-0.60

 Ashe

-0.60

POSITIVE LOGITS

 occasional

0.85

 disclaim

0.85

 caveats

0.70

 versa

0.69

 aforementioned

0.69

 obligatory

0.69

 usual

0.68

olis

0.65

zes

0.65

 modifiers

0.64

Activations Density 0.294%

entities or items listed in a series

references to different groups, individuals, and discussions related to various topics

No Comments

No Known Activations