INDEX

Explanations

things that are described as positive or beneficial

oai_token-act-pair · gpt-3.5-turbo

statements or phrases expressing a positive attribute or quality

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

iates

-0.74

iate

-0.66

ear

-0.64

 stalls

-0.63

 retrie

-0.61

 pursu

-0.59

 floats

-0.58

 Occup

-0.57

 registers

-0.57

ida

-0.57

POSITIVE LOGITS

senal

1.00

 undoubtedly

0.82

 indeed

0.79

ovie

0.73

ĻĤ

0.72

 nevertheless

0.72

KER

0.70

 nonetheless

0.70

 ALWAYS

0.68

Asset

0.67

Activations Density 0.292%

things that are described as positive or beneficial

statements or phrases expressing a positive attribute or quality

No Comments

No Known Activations