INDEX

Explanations

words starting with the letter "s" followed by other letters

oai_token-act-pair · gpt-3.5-turbo

references to the letter 's' in various contexts

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

hyde

-0.88

EStream

-0.73

 unrestricted

-0.69

 Wikileaks

-0.67

FINE

-0.67

tainment

-0.67

LEASE

-0.65

EStreamFrame

-0.62

 duplicate

-0.61

 incumbent

-0.61

POSITIVE LOGITS

arin

1.11

igm

1.11

arah

1.08

igma

1.06

idd

1.05

ams

1.02

outher

1.00

ixties

0.99

agging

0.98

uss

0.97

Activations Density 0.020%

words starting with the letter "s" followed by other letters

references to the letter 's' in various contexts

No Comments

No Known Activations