INDEX

Explanations

technical terms and bullet points in job descriptions or technical documentation

oai_token-act-pair · gpt-3.5-turbo

phrases indicating requirements or qualifications for roles or tasks

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Wander

-0.70

 cheers

-0.64

 madness

-0.62

 Gaal

-0.61

 Wrestle

-0.61

 taunt

-0.60

 charism

-0.59

 Tsukuyomi

-0.59

 Legend

-0.59

iously

-0.58

POSITIVE LOGITS

âĹı

1.30

âľ

1.22

âĢ¢

1.17

âĸł

1.08

Â·

1.04

âĪ

0.88

ethnic

0.84

âĨĳ

0.84

âĹ

0.84

âĺħ

0.82

Activations Density 0.113%

technical terms and bullet points in job descriptions or technical documentation

phrases indicating requirements or qualifications for roles or tasks

No Comments

No Known Activations