INDEX

Explanations

references to light emitting objects or phenomena, particularly those related to emergency situations

oai_token-act-pair · gpt-3.5-turbo

terms related to flares 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 @ 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

ointed

-0.80

rified

-0.75

phis

-0.74

nesota

-0.73

ournal

-0.71

ordan

-0.69

arnaev

-0.68

onut

-0.66

ivan

-0.65

æĸ¹

-0.65

POSITIVE LOGITS

 flares

1.26

 flare

1.13

 flared

1.03

eru

0.81

lights

0.81

othal

0.80

 lamps

0.79

fires

0.76

 capacitor

0.73

waters

0.73

Activations Density 0.020%

references to light emitting objects or phenomena, particularly those related to emergency situations

terms related to flares in various contexts

No Comments

No Known Activations