INDEX

Explanations

terms related to data encoding and decoding

oai_token-act-pair · gpt-3.5-turbo

terms related to data encoding and decoding processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

giving

-0.80

inness

-0.73

profit

-0.73

iability

-0.72

alsh

-0.70

zanne

-0.69

lv

-0.69

dash

-0.69

ivot

-0.68

lying

-0.67

POSITIVE LOGITS

 encoded

1.27

 encoding

1.19

 encode

1.12

oded

1.05

 codec

0.94

oding

0.92

ãĥĥãĥī

0.89

 decoding

0.86

 encrypted

0.84

 defe

0.84

Activations Density 0.009%

terms related to data encoding and decoding

terms related to data encoding and decoding processes

No Comments

No Known Activations