INDEX

Explanations

words related to computer programming and technology, specifically focusing on terms related to programming languages and software development

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 'p'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 outpatient

-0.60

 undrafted

-0.55

 freeing

-0.54

 referen

-0.53

 fracture

-0.53

 sciences

-0.52

 defeats

-0.52

 galaxy

-0.52

 Galaxy

-0.52

hiba

-0.51

POSITIVE LOGITS

3.55

pie

1.79

py

1.69

pa

1.65

pas

1.65

pi

1.64

1.59

pn

1.59

ps

1.55

pd

1.53

Activations Density 0.019%

words related to computer programming and technology, specifically focusing on terms related to programming languages and software development

occurrences of the letter 'p'

No Comments

No Known Activations