INDEX

Explanations

terms related to large-scale objects or operations

oai_token-act-pair · gpt-3.5-turbo

references to large-scale measurements or constructs

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

beat

-0.73

rique

-0.69

ãĤ¡

-0.68

vous

-0.64

 Biden

-0.63

mony

-0.61

nyder

-0.61

 Protocol

-0.60

issance

-0.60

forcer

-0.60

POSITIVE LOGITS

 bulky

0.85

 stature

0.83

 è£ıè

0.82

 lumber

0.78

 arsen

0.77

 quantities

0.73

 diameter

0.73

 conglomer

0.72

 contiguous

0.71

ikuman

0.70

Activations Density 0.279%

terms related to large-scale objects or operations

references to large-scale measurements or constructs

No Comments

No Known Activations