INDEX

Explanations

mentions of specific names or terms

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ad" in various contexts

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

 Purg

-0.70

 Impossible

-0.69

İĭ

-0.68

 sidx

-0.68

Ń·

-0.67

 shorth

-0.66

 Ragnarok

-0.64

ĸļ

-0.64

 pole

-0.64

 Nemesis

-0.61

POSITIVE LOGITS

venture

1.33

apter

1.16

vance

1.14

ventures

1.13

ad

1.08

rian

1.05

visory

1.04

vertis

1.03

mir

1.02

roid

1.00

Activations Density 0.021%

mentions of specific names or terms

instances of the word "ad" in various contexts

No Comments

No Known Activations