INDEX

Explanations

references to specific dates and events

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate changes or developments in a narrative context

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 âĵĺ

-0.75

 unless

-0.64

unless

-0.60

!:

-0.57

?:

-0.55

 regardless

-0.55

"/>

-0.55

 Nanto

-0.54

 PLEASE

-0.53

 whereas

-0.53

POSITIVE LOGITS

)).

0.65

bernatorial

0.61

OPA

0.60

©¶æ

0.60

gil

0.58

 hiatus

0.55

alysed

0.55

idelines

0.55

 earlier

0.54

 inception

0.54

Activations Density 1.213%

references to specific dates and events

phrases that indicate changes or developments in a narrative context

No Comments

No Known Activations