INDEX

Explanations

words related to scientific concepts or discussions

oai_token-act-pair · gpt-3.5-turbo

mentions of science

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Seasons

-0.75

âĢ¢âĢ¢

-0.70

leased

-0.67

 Arrows

-0.66

zik

-0.65

pora

-0.65

terior

-0.65

ometown

-0.64

 Shades

-0.64

ividual

-0.63

POSITIVE LOGITS

 fiction

1.24

 Fiction

1.03

 science

0.91

science

0.89

 sciences

0.84

 literacy

0.82

istry

0.81

craft

0.81

mong

0.81

fiction

0.78

Activations Density 0.030%

words related to scientific concepts or discussions

mentions of science

No Comments

No Known Activations