INDEX

Explanations

terms related to stockpiling, holding, or stash

oai_token-act-pair · gpt-3.5-turbo

references to stockpiles or reserves of items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Outs

-0.71

Adult

-0.70

Wheel

-0.69

boxing

-0.67

lit

-0.66

english

-0.65

obo

-0.64

ĪĴ

-0.63

adv

-0.62

 Johann

-0.62

POSITIVE LOGITS

 stockp

1.17

 stockpile

1.14

 stash

1.04

 holdings

1.02

 stocked

0.94

 shelves

0.94

enza

0.91

tery

0.82

 stocks

0.81

 amassed

0.80

Activations Density 0.029%

terms related to stockpiling, holding, or stash

references to stockpiles or reserves of items

No Comments

No Known Activations