INDEX

Explanations

references to specific individuals or names, particularly those ending with 'ai'

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "AI" 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 @ 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

din

-0.70

 Frie

-0.70

 Rebell

-0.70

lain

-0.68

stable

-0.67

 skelet

-0.66

ienced

-0.65

err

-0.65

oppy

-0.63

dies

-0.63

POSITIVE LOGITS

ju

1.08

jin

1.07

ya

0.96

ji

0.92

yah

0.90

uno

0.85

ples

0.84

yan

0.83

jah

0.82

wi

0.82

Activations Density 0.018%

references to specific individuals or names, particularly those ending with 'ai'

occurrences of the term "AI" in various contexts

No Comments

No Known Activations