INDEX

Explanations

references to the name "Hong" in various contexts

oai_token-act-pair · gpt-3.5-turbo

occurrences of the string "ong" 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

SPD

-0.75

 Goat

-0.69

è¦ļéĨĴ

-0.69

 Bucks

-0.65

 Avalanche

-0.64

======

-0.63

Pwr

-0.63

 Canary

-0.62

FG

-0.61

 NASCAR

-0.61

POSITIVE LOGITS

ong

1.33

ratulations

0.99

regate

0.97

jiang

0.94

ongs

0.90

ueless

0.89

ongo

0.88

mong

0.88

atana

0.87

Ĥ¬

0.86

Activations Density 0.009%

references to the name "Hong" in various contexts

occurrences of the string "ong" in various contexts

No Comments

No Known Activations