INDEX

Explanations

numbers related to dates, quantities, or rankings

oai_token-act-pair · gpt-3.5-turbo

numerical values, particularly those associated with specific identifiers or data points

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

awaru

-0.83

riel

-0.69

ificial

-0.67

ilic

-0.66

olas

-0.66

ilitating

-0.66

ilant

-0.66

opped

-0.65

romeda

-0.64

rolet

-0.64

POSITIVE LOGITS

ÃĹ

0.75

%,

0.73

Â¢

0.73

âĢ²

0.69

0.68

Â°

0.68

Â½

0.67

 Intermediate

0.67

0.66

%-

0.65

Activations Density 0.179%

numbers related to dates, quantities, or rankings

numerical values, particularly those associated with specific identifiers or data points

No Comments

No Known Activations