INDEX

Explanations

technical information about computer systems

oai_token-act-pair · gpt-3.5-turbo

negative point-values or indicators of problems in a technical context

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dickinson

-0.76

¿½

-0.75

Ń·

-0.71

ãĤ¡

-0.71

ãĤ¸

-0.69

ernel

-0.68

 wedd

-0.68

 citiz

-0.67

 incorpor

-0.67

ulhu

-0.66

POSITIVE LOGITS

sama

1.06

based

0.95

=-=-=-=-=-=-=-=-

0.94

_-

0.94

san

0.93

selection

0.92

dimensional

0.91

mediated

0.89

sized

0.88

sense

0.88

Activations Density 0.450%

technical information about computer systems

negative point-values or indicators of problems in a technical context

No Comments

No Known Activations