INDEX

Explanations

references to a specific entity named "Lo" at varying levels of adoration and specificity

oai_token-act-pair · gpt-3.5-turbo

references to the term "Lo" 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

manship

-0.85

itated

-0.79

pillar

-0.77

EMENT

-0.70

 Equality

-0.70

rations

-0.69

itates

-0.68

chell

-0.68

itating

-0.67

ITAL

-0.66

POSITIVE LOGITS

veland

1.26

zzi

1.13

fty

1.11

ppy

1.07

vers

1.05

zzle

1.03

aned

1.01

vel

0.98

ven

0.98

pper

0.98

Activations Density 0.027%

references to a specific entity named "Lo" at varying levels of adoration and specificity

references to the term "Lo" in various contexts

No Comments

No Known Activations