INDEX

Explanations

references to the mineral "Ore"

oai_token-act-pair · gpt-3.5-turbo

references to the word 'Ore' in various contexts

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

":["

-0.72

BILITIES

-0.67

Applications

-0.64

Strong

-0.62

WAYS

-0.61

Traditional

-0.60

 Laos

-0.59

 Shepard

-0.59

 Patient

-0.59

--------------------------------------------------------

-0.58

POSITIVE LOGITS

heim

0.95

heimer

0.92

Ore

0.91

gan

0.91

ore

0.83

asure

0.82

gon

0.79

oin

0.78

zan

0.77

ghan

0.75

Activations Density 0.014%

references to the mineral "Ore"

references to the word 'Ore' in various contexts

No Comments

No Known Activations