INDEX

Explanations

positive comments or thoughts

oai_token-act-pair · gpt-3.5-turbo

statements of thought or opinion expressed about various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

*/(

-0.82

ESE

-0.82

APD

-0.76

aptic

-0.75

soever

-0.72

-0.71

ICLE

-0.71

emphasis

-0.69

umbnails

-0.68

vale

-0.67

POSITIVE LOGITS

 hilarious

1.07

 funny

1.04

 joking

1.04

 nuts

1.04

 crazy

1.03

 gonna

1.03

 cool

1.03

 cute

1.02

 invincible

0.98

 alright

0.95

Activations Density 0.163%

positive comments or thoughts

statements of thought or opinion expressed about various subjects

No Comments

No Known Activations