INDEX

Explanations

nationalities

oai_token-act-pair · gpt-3.5-turbo

plurals and demonyms related to various ethnic and national groups

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥ¼ãĥ³

-0.76

-+-+

-0.73

LET

-0.69

SET

-0.67

shown

-0.67

Performance

-0.66

 CASE

-0.65

é¾

-0.65

inventoryQuantity

-0.65

 srfAttach

-0.64

POSITIVE LOGITS

ervatives

1.07

ktop

0.90

aurus

0.88

hip

0.88

who

0.84

paces

0.84

folk

0.83

'/

0.80

rats

0.79

 residing

0.79

Activations Density 0.152%

nationalities

plurals and demonyms related to various ethnic and national groups

No Comments

No Known Activations