INDEX

Explanations

numbers and measurements

oai_token-act-pair · gpt-3.5-turbo

numerical values related to time, measurements, and quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ashtra

-0.67

ulkan

-0.66

ande

-0.62

wart

-0.61

urdy

-0.60

 reeling

-0.60

raught

-0.59

doing

-0.58

products

-0.58

 sorts

-0.58

POSITIVE LOGITS

 increments

1.09

 apiece

0.95

bsp

0.82

undred

0.82

ousand

0.74

Ago

0.69

 intervals

0.67

 teasp

0.63

è£

0.63

GHz

0.63

Activations Density 0.172%

numbers and measurements

numerical values related to time, measurements, and quantities

No Comments

No Known Activations