INDEX

Explanations

adjectives describing a range of qualities from positive to negative

oai_token-act-pair · gpt-3.5-turbo

comparisons involving extremes, particularly in moral or ethical contexts

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

pta

-0.90

regation

-0.86

glers

-0.85

olphins

-0.84

EStream

-0.81

 Flavoring

-0.80

ortmund

-0.78

Ws

-0.77

anwhile

-0.77

onds

-0.76

POSITIVE LOGITS

 downright

1.17

 incomprehensible

1.09

 indistinguishable

1.08

 resistant

1.07

 impossible

1.06

 impractical

1.06

 lacking

1.03

 profitable

1.02

 agreeable

1.02

 attractive

1.01

Activations Density 0.477%

adjectives describing a range of qualities from positive to negative

comparisons involving extremes, particularly in moral or ethical contexts

No Comments

No Known Activations

adjectives describing a range of qualities from positive to negative

comparisons involving extremes, particularly in moral or ethical contexts

No Comments

No Known Activations