INDEX

Explanations

phrases related to the process of learning new skills or knowledge

oai_token-act-pair · gpt-3.5-turbo

infinitive verb phrases that indicate actions or processes of learning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

lights

-0.85

listed

-0.76

teen

-0.76

isite

-0.73

 Appears

-0.66

holder

-0.66

 Hartford

-0.66

nces

-0.65

termination

-0.65

quished

-0.65

POSITIVE LOGITS

 differentiate

1.23

 cope

1.22

 recognize

1.22

 behave

1.20

 manipulate

1.14

 wrestle

1.13

 tolerate

1.13

 distinguish

1.08

 recognise

1.08

 regulate

1.04

Activations Density 0.110%

phrases related to the process of learning new skills or knowledge

infinitive verb phrases that indicate actions or processes of learning

No Comments

No Known Activations