INDEX

Explanations

references to rankings or ordering within a specific context

oai_token-act-pair · gpt-3.5-turbo

references to rankings or statistical measurements related to a group of items or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

assetsadobe

-0.82

ahime

-0.78

atown

-0.69

rush

-0.65

heit

-0.63

gue

-0.61

nep

-0.60

 Decoder

-0.59

inventoryQuantity

-0.59

Allows

-0.59

POSITIVE LOGITS

ses

1.11

 aforementioned

1.08

 largest

0.97

 biggest

0.96

 brightest

0.91

 dozen

0.89

 remaining

0.89

 nine

0.86

 richest

0.85

 finalists

0.85

Activations Density 0.096%

references to rankings or ordering within a specific context

references to rankings or statistical measurements related to a group of items or entities

No Comments

No Known Activations