INDEX

Explanations

validity or the concept of something being valid

oai_token-act-pair · gpt-3.5-turbo

references to the concept of validity in various contexts

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

ILA

-0.79

hedon

-0.76

 Sisters

-0.72

Downloadha

-0.68

 Carbuncle

-0.64

RAW

-0.64

cipl

-0.63

hani

-0.63

peak

-0.61

 Plants

-0.61

POSITIVE LOGITS

 valid

1.19

ating

1.00

 validity

0.92

Valid

0.89

 Valid

0.88

valid

0.88

ators

0.86

ational

0.82

ations

0.81

atable

0.80

Activations Density 0.004%

validity or the concept of something being valid

references to the concept of validity in various contexts

No Comments

No Known Activations