INDEX

Explanations

references to the year 2016

oai_token-act-pair · gpt-3.5-turbo

occurrences of the year 2016

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

bor

-0.85

¿½

-0.75

angs

-0.73

ospital

-0.72

 encl

-0.71

aith

-0.70

warts

-0.69

bors

-0.69

 tremend

-0.69

semb

-0.67

POSITIVE LOGITS

0.96

0.95

å¹

0.93

0.86

0.81

 Countdown

0.77

0.76

0.73

âĢ²

0.72

âĶĢâĶĢâĶĢâĶĢ

0.71

Activations Density 0.025%

references to the year 2016

occurrences of the year 2016

No Comments

No Known Activations