INDEX

Explanations

instances of someone escaping or avoiding harm

oai_token-act-pair · gpt-3.5-turbo

language related to escape or liberation

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

soDeliveryDate

-0.89

 Compass

-0.70

MX

-0.66

Ped

-0.66

paper

-0.64

yi

-0.64

 Huawei

-0.63

 indicators

-0.60

cientious

-0.60

 Associ

-0.59

POSITIVE LOGITS

 confines

0.89

 boredom

0.85

 clut

0.78

 gunfire

0.76

 bounds

0.76

 altogether

0.75

 fumes

0.74

 wilderness

0.71

 obscurity

0.71

 tears

0.68

Activations Density 0.209%

instances of someone escaping or avoiding harm

language related to escape or liberation

No Comments

No Known Activations