INDEX

Explanations

words related to rules, regulations, and instructions

oai_token-act-pair · gpt-3.5-turbo

words and phrases related to prescriptive instructions or recommendations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 shine

-0.62

 Spiegel

-0.62

 pulp

-0.61

 fracturing

-0.61

 scandal

-0.61

 crotch

-0.60

 marrow

-0.59

 disturb

-0.59

 reckoning

-0.59

 grief

-0.57

POSITIVE LOGITS

orial

0.79

IAL

0.76

ums

0.76

ially

0.76

ual

0.76

esses

0.75

ciating

0.75

ocument

0.74

ãĤ·ãĥ£

0.74

Ã©t

0.74

Activations Density 0.121%

words related to rules, regulations, and instructions

words and phrases related to prescriptive instructions or recommendations

No Comments

No Known Activations