INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

occurrences of variations of the word "ind."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Fired

-0.62

 guidance

-0.60

SPL

-0.59

 halfway

-0.58

 litter

-0.57

 gorilla

-0.57

TION

-0.57

sonian

-0.56

 waivers

-0.56

 deaf

-0.56

POSITIVE LOGITS

ividually

1.49

ebted

1.39

irect

1.39

epend

1.38

etermin

1.38

igenous

1.33

ignant

1.30

oor

1.30

icative

1.28

icator

1.28

Activations Density 0.015%

proper nouns

occurrences of variations of the word "ind."

No Comments

No Known Activations