INDEX

Explanations

phrases indicating finality or conclusion

oai_token-act-pair · gpt-3.5-turbo

words that indicate possession or relation

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

rul

-0.82

 destro

-0.73

 nodd

-0.69

ahon

-0.68

 seiz

-0.67

 cous

-0.66

ãĥ¼ãĥĨãĤ£

-0.64

 Citiz

-0.62

 warr

-0.60

Reloaded

-0.59

POSITIVE LOGITS

 lest

0.93

 Â©

0.89

.:

0.84

 when

0.77

.#

0.76

.</

0.75

.[

0.74

.<

0.74

Â©

0.74

Activations Density 0.463%

phrases indicating finality or conclusion

words that indicate possession or relation

No Comments

No Known Activations