INDEX

Explanations

words or phrases that suggest or imply something

oai_token-act-pair · gpt-3.5-turbo

phrases that suggest indirect meaning or insinuations

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

frey

-0.81

mir

-0.80

afa

-0.79

Bio

-0.76

agn

-0.75

acking

-0.74

ICAN

-0.72

unker

-0.72

fc

-0.71

tp

-0.68

POSITIVE LOGITS

 implied

1.32

 imply

0.93

 guiActiveUn

0.91

 implication

0.86

 warranties

0.84

 inference

0.82

 warnings

0.81

ratulations

0.81

 implicitly

0.79

 insin

0.78

Activations Density 0.008%

words or phrases that suggest or imply something

phrases that suggest indirect meaning or insinuations

No Comments

No Known Activations