INDEX

Explanations

instances where data is stored or retrievable

oai_token-act-pair · gpt-3.5-turbo

terms related to data storage and memory

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eds

-0.70

nir

-0.69

 Ascend

-0.66

aldi

-0.66

sky

-0.65

naire

-0.65

lez

-0.64

SHIP

-0.64

 Cosponsors

-0.64

doms

-0.63

POSITIVE LOGITS

 storing

1.01

ãĤª

0.85

 secrets

0.84

 stored

0.79

ãĤ´

0.78

omach

0.77

 potatoes

0.77

 shelves

0.74

IER

0.73

 storage

0.73

Activations Density 0.027%

instances where data is stored or retrievable

terms related to data storage and memory

No Comments

No Known Activations