INDEX

Explanations

phrases related to completion or resolution

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing the concept of "out of the way."

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

ivot

-0.72

IELD

-0.71

illac

-0.69

olo

-0.67

LI

-0.67

ãĥį

-0.67

raud

-0.66

olicy

-0.65

owski

-0.65

owicz

-0.64

POSITIVE LOGITS

 gate

1.04

box

1.01

 frying

0.99

 closet

0.93

 gates

0.93

 equation

0.88

 loop

0.86

 woods

0.85

way

0.84

 womb

0.82

Activations Density 0.074%

phrases related to completion or resolution

phrases emphasizing the concept of "out of the way."

No Comments

No Known Activations