INDEX

Explanations

names of individuals, potentially authors or public figures

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or notable figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

vous

-0.75

ez

-0.73

 reflex

-0.71

otle

-0.70

vez

-0.67

crew

-0.61

due

-0.60

orius

-0.59

ulic

-0.59

Lab

-0.58

POSITIVE LOGITS

owship

1.20

atio

0.86

uates

0.76

 Beckham

0.74

iquette

0.73

vu

0.72

berra

0.72

ername

0.71

insk

0.70

ibur

0.70

Activations Density 0.054%

names of individuals, potentially authors or public figures

references to specific individuals or notable figures

No Comments

No Known Activations