INDEX

Explanations

words related to something being recognized or acknowledged

oai_token-act-pair · gpt-3.5-turbo

the phrase "have been known" 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 @ 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

otion

-0.77

aminer

-0.77

rentice

-0.75

erva

-0.70

plet

-0.69

anco

-0.67

ovie

-0.66

reau

-0.66

boa

-0.66

olicy

-0.65

POSITIVE LOGITS

lege

0.89

Ð»

0.86

ledged

0.85

cut

0.78

ÙĨ

0.70

ãĥĥãĥĪ

0.69

saf

0.69

 quantity

0.68

ãĤ±

0.67

sie

0.67

Activations Density 0.037%

words related to something being recognized or acknowledged

the phrase "have been known" in various contexts

No Comments

No Known Activations