INDEX

Explanations

terms related to being outdated or obsolete

oai_token-act-pair · gpt-3.5-turbo

terms related to obsolescence and outdatedness

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 sensation

-0.67

ather

-0.67

 Salman

-0.64

 Participants

-0.63

tu

-0.61

pled

-0.60

pez

-0.59

 Gabriel

-0.58

sum

-0.58

 Booker

-0.58

POSITIVE LOGITS

 obsolete

3.49

 outdated

3.20

 archaic

2.09

 antiqu

2.06

 defunct

2.00

solete

1.85

 extinct

1.77

 deprecated

1.73

 redundant

1.65

 stale

1.61

Activations Density 0.019%

terms related to being outdated or obsolete

terms related to obsolescence and outdatedness

No Known Activations