INDEX

Explanations

years mentioned in a document

oai_token-act-pair · gpt-3.5-turbo

mentions of specific years or dates

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 invoking

-0.72

 administr

-0.72

hops

-0.71

phy

-0.70

alls

-0.68

 hungry

-0.67

estine

-0.66

rosis

-0.66

enance

-0.66

ifully

-0.66

POSITIVE LOGITS

å¹

1.32

 STATS

0.97

 Ranked

0.93

 Salary

0.89

 edition

0.88

 Annual

0.87

WCS

0.86

 Chevrolet

0.85

 INTER

0.85

 ELECT

0.83

Activations Density 0.068%

years mentioned in a document

mentions of specific years or dates

No Comments

No Known Activations