INDEX

Explanations

text related to events or actions that happened before a specific point in time

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "prior" and its variants indicating past events or conditions

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

rosso

-0.75

RO

-0.66

aden

-0.66

_-

-0.65

Hub

-0.64

Baby

-0.63

VILLE

-0.62

beans

-0.60

cn

-0.59

stem

-0.58

POSITIVE LOGITS

itiz

1.48

ities

1.35

itized

1.25

etheless

0.93

ITIES

0.92

ITY

0.92

ity

0.84

izations

0.83

lly

0.82

itatively

0.82

Activations Density 0.036%

text related to events or actions that happened before a specific point in time

occurrences of the word "prior" and its variants indicating past events or conditions

No Comments

No Known Activations