INDEX

Explanations

prepositional phrases indicating physical entry into a location

oai_token-act-pair · gpt-3.5-turbo

occurrences of the phrase "walked into."

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

entimes

-0.76

 timetable

-0.71

Tips

-0.67

gently

-0.65

ice

-0.63

nah

-0.62

umably

-0.62

anooga

-0.62

 tongue

-0.62

culosis

-0.62

POSITIVE LOGITS

 Valhalla

0.83

 adulthood

0.82

 orbit

0.76

jection

0.73

 existence

0.69

jected

0.64

veyard

0.64

 Slot

0.63

 hypers

0.63

 parliament

0.63

Activations Density 0.079%

prepositional phrases indicating physical entry into a location

occurrences of the phrase "walked into."

No Comments

No Known Activations