INDEX

Explanations

new technology, products, or concepts that have been developed or created

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "developed."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

que

-0.81

 Keeper

-0.79

lua

-0.73

knife

-0.70

wa

-0.70

pir

-0.69

wait

-0.68

mir

-0.67

nee

-0.67

ipe

-0.66

POSITIVE LOGITS

developed

1.07

 developed

1.02

 develops

0.92

 developing

0.88

urally

0.82

 behavi

0.79

 Seym

0.79

irtual

0.78

Ô

0.76

icter

0.76

Activations Density 0.011%

new technology, products, or concepts that have been developed or created

occurrences of the word "developed."

No Comments

No Known Activations