INDEX

Explanations

dates and names related to locations

oai_token-act-pair · gpt-3.5-turbo

the repeated mention of a specific name, likely a key individual or entity related to the context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

kefeller

-0.66

iculty

-0.66

 heels

-0.66

ugu

-0.65

 convention

-0.64

raltar

-0.64

ORD

-0.64

USS

-0.64

 stride

-0.63

 Duff

-0.62

POSITIVE LOGITS

nets

1.11

Son

0.95

oma

0.95

ners

0.95

 Gohan

0.95

hesis

0.93

nen

0.90

heses

0.90

ata

0.90

Son

0.88

Activations Density 0.011%

dates and names related to locations

the repeated mention of a specific name, likely a key individual or entity related to the context

No Comments

No Known Activations