INDEX

Explanations

words related to personal names, possibly focusing on the name "Ja" in particular

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of a specific entity or term, particularly focusing on "ja"

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

ships

-0.93

sworth

-0.90

holders

-0.77

sheet

-0.77

balance

-0.73

drawn

-0.72

mate

-0.71

tons

-0.71

 constit

-0.69

sheets

-0.69

POSITIVE LOGITS

quez

1.13

ques

0.91

udeau

0.90

Å¡

0.86

Ã±

0.86

Ã°

0.86

vel

0.83

zyk

0.81

Äĩ

0.81

uthor

0.81

Activations Density 0.013%

words related to personal names, possibly focusing on the name "Ja" in particular

repeated mentions of a specific entity or term, particularly focusing on "ja"

No Comments

No Known Activations