INDEX

Explanations

positive feedback or praise

oai_token-act-pair · gpt-3.5-turbo

instances of the word "rave" and related emotional descriptors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

lr

-0.80

nel

-0.78

esville

-0.75

rast

-0.72

nect

-0.71

oglu

-0.68

uyomi

-0.68

auga

-0.66

ratom

-0.64

alities

-0.64

POSITIVE LOGITS

 rave

1.14

 reviews

0.81

ochet

0.77

 praise

0.73

uve

0.73

 glowing

0.69

ranging

0.69

 farewell

0.66

EVA

0.64

âĺħâĺħ

0.64

Activations Density 0.022%

positive feedback or praise

instances of the word "rave" and related emotional descriptors

No Comments

No Known Activations