INDEX

Explanations

specific phrases with the letters "af" in them

oai_token-act-pair · gpt-3.5-turbo

instances of the prefix "af," often indicating a focus on conditions or states of being

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

ANGE

-0.73

MER

-0.73

Accessory

-0.71

URES

-0.69

 Mothers

-0.68

DOWN

-0.67

Water

-0.66

 Rubin

-0.66

Med

-0.65

PATH

-0.64

POSITIVE LOGITS

icion

1.41

rican

1.29

rica

1.09

avorite

1.05

riad

1.00

riend

0.93

eatures

0.91

ghan

0.88

af

0.87

ird

0.85

Activations Density 0.003%

specific phrases with the letters "af" in them

instances of the prefix "af," often indicating a focus on conditions or states of being

No Comments

No Known Activations