INDEX

Explanations

phrases related to specific sections or parts within a larger entity

oai_token-act-pair · gpt-3.5-turbo

the term "segment" and its variations, indicating a focus on structured parts or divisions within a context

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

 opio

-0.72

 lett

-0.67

 mother

-0.64

 fists

-0.62

 realism

-0.62

dra

-0.61

stroke

-0.60

ggies

-0.60

âĺħâĺħ

-0.59

CHA

-0.59

POSITIVE LOGITS

ation

1.06

ated

0.88

naire

0.83

ed

0.81

eston

0.80

regate

0.80

ational

0.80

arily

0.79

meal

0.79

ials

0.79

Activations Density 0.012%

phrases related to specific sections or parts within a larger entity

the term "segment" and its variations, indicating a focus on structured parts or divisions within a context

No Comments

No Known Activations